{"$update": {"344": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "345": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "346": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "347": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "340": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "341": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "342": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "343": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "348": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "349": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1653": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2317": {"$update": {"model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2310": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "298": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "299": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "296": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "297": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "294": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "295": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "292": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "293": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "290": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "291": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "270": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "271": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "272": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "273": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "274": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "275": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "276": {"$update": {"model_description": "This model detects protein overexpression based on the presence of mutations.The detection of the protein without an associated mutation indicates that the protein is likely to be expressed at low or basal levels. The detection of the protein with the mutation indicates that the protein is likely overexpressed. This model reflects how certain proteins are functional with and without mutations. For example, efflux pump subunits and regulators are functional with mutations and without mutations. Without mutations, efflux pump subunits and regulators are usually expressed at a low level. When an efflux pump regulator has a mutation, it can cause the overexpression of the efflux pump it is responsible for regulating, leading to resistance to specific drugs. Protein overexpression models have two parameters: a curated BLASTP cutoff, and a curated set of mutations (single resistance variants, frameshift mutations, indels, etc.) shown clinically to confer resistance. This model type is a combination of the protein homolog and protein variant model. A detected hit can be categorized as Perfect, Strict, or Loose with no mutation(s) or as Strict or Loose with mutation(s).", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "277": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "278": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "279": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1132": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2262": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2260": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2261": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2267": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "ARO_category": {"$update": {"40420": {"$update": {"category_aro_name": "determinant of nitrofuran resistance"}}}}, "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}, "snp": {"$update": {"param_value": {"$insert": {"8195": "R133S"}}, "clinical": {"$insert": {"8195": "R133S"}}}}}, "$insert": {"40394": {"param_value": {"8192": "Q44STOP", "8193": "K141STOP", "8194": "E233STOP"}, "param_type_id": "40394", "param_type": "nonsense mutation", "param_description": "A nucleotide substitution resulting in a change from an amino acid codon to a STOP codon. Nonsense mutations truncate protein translation prematurely, resulting in a defective or completely inactive protein. In CARD, nonsense mutations may be attached to models using the notation: [wild type amino acid][position][STOP] (e.g. Q42STOP). This parameter is not currently used in detection algorithms."}, "41343": {"param_value": {"8190": "-nt603:C", "8189": "-nt25:T"}, "param_type_id": "41343", "param_type": "deletion mutation from nucleotide sequence", "param_description": "A subtype of the deletion mutation detection model parameter. This parameter is used when a set of deletion mutations is reported in a nucleotide sequence format. Such mutations may be of variable length - possibly causing a frameshift, but not premature termination of functional knockout. Mutation parameters of this type are reported in the CARD with the notation: [-]nt[position]:[nucleotides]."}, "41342": {"param_value": {"8187": "-MTPTIELICGHRSIRHFTDEPISEAQ1-26", "8188": "-QYDEQLA191-197"}, "param_type_id": "41342", "param_type": "deletion mutation from peptide sequence", "param_description": "A subtype of the deletion mutation detection model parameter. This parameter is used when a set of deletion mutations is reported in a peptide sequence format. These are specific to codon deletions, where a multiple of 3 nucleotides are deleted. Mutations of this type are reported in the CARD with the notation: [-][AAs][position range]."}}}}}, "2264": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2265": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2445": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "108": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "109": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "102": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "103": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "100": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}, "$delete": ["40334"], "$insert": {"41345": {"param_value": {"8091": "+nt811:1", "8090": "+nt338:A"}, "param_type_id": "41345", "param_type": "insertion mutation from nucleotide sequence", "param_description": "A subtype of the insertion mutation detection model parameter. This parameter is used when a set of insertion mutations is reported in a nucleotide sequence format. Such mutations may be of variable length - possibly causing a frameshift, but not causing premature termination or a functional knockout. Mutation parameters of this type are reported in CARD with the notation: [+]nt[position]:[nucleotides]."}, "41343": {"param_value": {"8088": "-nt768:G", "8094": "-nt65:1", "8093": "-nt703:T", "8089": "-nt110:A", "8087": "-nt1290:C"}, "param_type_id": "41343", "param_type": "deletion mutation from nucleotide sequence", "param_description": "A subtype of the deletion mutation detection model parameter. This parameter is used when a set of deletion mutations is reported in a nucleotide sequence format. Such mutations may be of variable length - possibly causing a frameshift, but not premature termination of functional knockout. Mutation parameters of this type are reported in the CARD with the notation: [-]nt[position]:[nucleotides]."}}}}}, "101": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "106": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "107": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "104": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "105": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2046": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2047": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2044": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2045": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2042": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2043": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2040": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2041": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2048": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2049": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1213": {"$update": {"model_description": "This model detects protein overexpression based on the presence of mutations.The detection of the protein without an associated mutation indicates that the protein is likely to be expressed at low or basal levels. The detection of the protein with the mutation indicates that the protein is likely overexpressed. This model reflects how certain proteins are functional with and without mutations. For example, efflux pump subunits and regulators are functional with mutations and without mutations. Without mutations, efflux pump subunits and regulators are usually expressed at a low level. When an efflux pump regulator has a mutation, it can cause the overexpression of the efflux pump it is responsible for regulating, leading to resistance to specific drugs. Protein overexpression models have two parameters: a curated BLASTP cutoff, and a curated set of mutations (single resistance variants, frameshift mutations, indels, etc.) shown clinically to confer resistance. This model type is a combination of the protein homolog and protein variant model. A detected hit can be categorized as Perfect, Strict, or Loose with no mutation(s) or as Strict or Loose with mutation(s).", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "375", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}, "$delete": ["40334"], "$insert": {"41345": {"param_value": {"8153": "+nt410:TGTTCATCGAACTCTGCGAGCAG"}, "param_type_id": "41345", "param_type": "insertion mutation from nucleotide sequence", "param_description": "A subtype of the insertion mutation detection model parameter. This parameter is used when a set of insertion mutations is reported in a nucleotide sequence format. Such mutations may be of variable length - possibly causing a frameshift, but not causing premature termination or a functional knockout. Mutation parameters of this type are reported in CARD with the notation: [+]nt[position]:[nucleotides]."}}}}}, "2038": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1210": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2688": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2689": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "ARO_category": {"$insert": {"40403": {"category_aro_name": "determinant of oxazolidinone resistance", "category_aro_cvterm_id": "40403", "category_aro_accession": "3003747", "category_aro_description": "Enzymes, other proteins or other gene products shown clinically to confer resistance to oxazolidinone (ie., linezolid) antibiotics."}}}, "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}, "snp": {"$update": {"param_value": {"7940": "C2579T", "7941": "G2604T"}, "clinical": {"7940": "C2579T", "7941": "G2604T"}}}}}}}, "2685": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2686": {"$update": {"model_param": {"$update": {"41141": {"$update": {"param_description": "This detection model parameter describes efflux pump components that are to be detected together (e.g., efflux pump subunits and regulators) using sequential model IDs, separated by commas. For example: 2685,440,1925,1305."}}}}}}, "2680": {"$update": {"model_param": {"$update": {"41141": {"$update": {"param_description": "This detection model parameter describes efflux pump components that are to be detected together (e.g., efflux pump subunits and regulators) using sequential model IDs, separated by commas. For example: 2685,440,1925,1305."}}}}}}, "2681": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2682": {"$update": {"model_param": {"$update": {"41141": {"$update": {"param_description": "This detection model parameter describes efflux pump components that are to be detected together (e.g., efflux pump subunits and regulators) using sequential model IDs, separated by commas. For example: 2685,440,1925,1305."}}}}}}, "2683": {"$update": {"model_param": {"$update": {"41141": {"$update": {"param_description": "This detection model parameter describes efflux pump components that are to be detected together (e.g., efflux pump subunits and regulators) using sequential model IDs, separated by commas. For example: 2685,440,1925,1305."}}}}}}, "99": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "98": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "91": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "450", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "90": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "93": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "92": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "95": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "94": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "97": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "96": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1623": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1622": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1621": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1620": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1627": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1626": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1625": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1624": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1999": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1998": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1629": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1628": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "559": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "558": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "555": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "554": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "557": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "556": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "551": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "550": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "553": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "552": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1502": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1439": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1199": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1198": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1191": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "700", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1190": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1193": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1192": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1195": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1194": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1197": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1196": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1759": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1758": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1757": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "675", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1756": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1755": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1754": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1753": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1752": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1751": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1750": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1177": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1176": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}, "40394": {"$update": {"param_type": "nonsense mutation", "param_description": "A nucleotide substitution resulting in a change from an amino acid codon to a STOP codon. Nonsense mutations truncate protein translation prematurely, resulting in a defective or completely inactive protein. In CARD, nonsense mutations may be attached to models using the notation: [wild type amino acid][position][STOP] (e.g. Q42STOP). This parameter is not currently used in detection algorithms."}}}, "$delete": ["40334"], "$insert": {"41345": {"param_value": {"8048": "+nt1365:G", "8066": "+nt325:T", "8051": "+nt392:T", "8068": "+nt135:T", "8067": "+nt518:GGTC"}, "param_type_id": "41345", "param_type": "insertion mutation from nucleotide sequence", "param_description": "A subtype of the insertion mutation detection model parameter. This parameter is used when a set of insertion mutations is reported in a nucleotide sequence format. Such mutations may be of variable length - possibly causing a frameshift, but not causing premature termination or a functional knockout. Mutation parameters of this type are reported in CARD with the notation: [+]nt[position]:[nucleotides]."}, "41343": {"param_value": {"8040": "-nt571:GGCGGC", "8041": "-nt1501:G", "8043": "-nt1525:A", "8045": "-nt1293:G", "8055": "-nt126:G", "8069": "-nt241:G", "8070": "-nt60:A", "8054": "-nt368:G", "8065": "-nt54:C", "8053": "-nt249:G", "8056": "-nt81:C"}, "param_type_id": "41343", "param_type": "deletion mutation from nucleotide sequence", "param_description": "A subtype of the deletion mutation detection model parameter. This parameter is used when a set of deletion mutations is reported in a nucleotide sequence format. Such mutations may be of variable length - possibly causing a frameshift, but not premature termination of functional knockout. Mutation parameters of this type are reported in the CARD with the notation: [-]nt[position]:[nucleotides]."}}}}}, "1175": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}, "$delete": ["40334"], "$insert": {"41344": {"param_value": {"3851": "+MPL110-112"}, "param_type_id": "41344", "param_type": "insertion mutation from peptide sequence", "param_description": "A subtype of the insertion mutation detection model parameter. This parameter is used when a set of insertion mutations is reported in a peptide sequence format. These are specific to codon insertions, where a multiple of three nucleotides are inserted. This does not cause a frameshift mutation. Mutation parameters of this type are reported in CARD with the notation: [+][AAs][position range]."}}}}}, "1174": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1173": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1172": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1171": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1170": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1179": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1178": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "511": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "510": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1005": {"$update": {"model_description": "This model detects protein overexpression based on the presence of mutations.The detection of the protein without an associated mutation indicates that the protein is likely to be expressed at low or basal levels. The detection of the protein with the mutation indicates that the protein is likely overexpressed. This model reflects how certain proteins are functional with and without mutations. For example, efflux pump subunits and regulators are functional with mutations and without mutations. Without mutations, efflux pump subunits and regulators are usually expressed at a low level. When an efflux pump regulator has a mutation, it can cause the overexpression of the efflux pump it is responsible for regulating, leading to resistance to specific drugs. Protein overexpression models have two parameters: a curated BLASTP cutoff, and a curated set of mutations (single resistance variants, frameshift mutations, indels, etc.) shown clinically to confer resistance. This model type is a combination of the protein homolog and protein variant model. A detected hit can be categorized as Perfect, Strict, or Loose with no mutation(s) or as Strict or Loose with mutation(s).", "model_param": {"$update": {"40494": {"$update": {"param_type": "frameshift mutation", "param_description": "A frameshift is a type of genetic mutation caused by a nucleotide insertion or deletion \u2260 3 bases. This changes the grouping of codons and thus the reading frame during translation, resulting in a incomplete or inactive protein product. Many frameshift mutations generate downstream STOP codons, resulting in premature peptide translation termination. Frameshifts may also confer antibiotic resistance through partial or total protein loss-of-function. Frameshift mutations are included with relevant models when applicable, with the following notation: [wild-type AA][position]fs;[[wild-type AA][position]STOP], where AA is an amino acid. If the premature STOP codon position is unknown or does not exist, [wild-type AA][position]fs is sufficient. This parameter is currently not included in detection algorithms."}}, "blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}, "$delete": ["40334"], "$insert": {"41343": {"param_value": {"8025": "-nt1130:2"}, "param_type_id": "41343", "param_type": "deletion mutation from nucleotide sequence", "param_description": "A subtype of the deletion mutation detection model parameter. This parameter is used when a set of deletion mutations is reported in a nucleotide sequence format. Such mutations may be of variable length - possibly causing a frameshift, but not premature termination of functional knockout. Mutation parameters of this type are reported in the CARD with the notation: [-]nt[position]:[nucleotides]."}, "41342": {"param_value": {"3893": "-S128"}, "param_type_id": "41342", "param_type": "deletion mutation from peptide sequence", "param_description": "A subtype of the deletion mutation detection model parameter. This parameter is used when a set of deletion mutations is reported in a peptide sequence format. These are specific to codon deletions, where a multiple of 3 nucleotides are deleted. Mutations of this type are reported in the CARD with the notation: [-][AAs][position range]."}}}}}, "1285": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1284": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1287": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "512": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1281": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1280": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1283": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1282": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1003": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1289": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1288": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "514": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1579": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1578": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "689": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "688": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "685": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "684": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "687": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "686": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "681": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "680": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "683": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "682": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1227": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1226": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1240": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "621": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "873": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1224": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "627": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1222": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_name": "fosA", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1221": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "624": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2743": {"$update": {"model_param": {"$update": {"41141": {"$update": {"param_description": "This detection model parameter describes efflux pump components that are to be detected together (e.g., efflux pump subunits and regulators) using sequential model IDs, separated by commas. For example: 2685,440,1925,1305."}}}}}}, "407": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1370": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "405": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1372": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1375": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1374": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1377": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "400": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1379": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1378": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1342": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "409": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "408": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "453": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2747": {"$update": {"ARO_description": "AcrEF-TolC efflux pump system of E. coli confers resistance to fluoroquinolones (ciprofloxacin).", "ARO_name": "AcrEF-TolC confers resistance to ciprofloxacin", "model_param": {"$update": {"41141": {"$update": {"param_description": "This detection model parameter describes efflux pump components that are to be detected together (e.g., efflux pump subunits and regulators) using sequential model IDs, separated by commas. For example: 2685,440,1925,1305."}}}}}}, "454": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2746": {"$update": {"model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1345": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2749": {"$update": {"ARO_description": "lnuG is a transposon-mediated lincosamide nucleotidyltransferase found in Enterococcus faecalis on Tn6260.", "model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_sequences": {"$update": {"sequence": {"4181": {"dna_sequence": {"fmax": "4615", "fmin": "3811", "accession": "KX470419.1", "strand": "+", "sequence": "TTGTTAAAACAAAAGGAACTAATGGCAAGGGTTAAGGAACTTGTCCAGTCAGATGAACGAATATCTGCTTGTATGATGTATGGCTCTTTTACAAAAGGAGAGGGAGATCAATACTCTGATATAGAATATTACGTTTTTCTAAAAGATGATACAATTTCCACCTTTGATTCAGCAAAATGGCTAAATGAAGTCGCTTCCTACACTTTACTCTATCAAAATGAGTACGGTACGGAAGTAGTAATTTTTGAAAATCTAATACGTGGTGAATTTCATTTCCTTTCCGAAAACGAAATGAATATTATTCCTTCATTCAAAGAATCAGGCTACATTCCTGACACAAAAGCAATGTTTATTTATGATGAAACAGGACAACTAGAATTGTATTTATCAGAGTTGGAAGGTCCGGGACCAAATAGACTTACAGAAGAAAACGTAAATTTTTTATTGAATAATTTTTCCAACCTATGGTTAATGGGGATTAATGTTCTTAAAAGAGGGGAAAATGCACGTTCACTGGAACTTTTATCTCAATTACAAAAAAATATACTACAACTCATTCGAATTGCGGAAGAAAATGCCGATAATTGGTTTAATATGACAAAGAATCTTGAAAAAGAAATTAGTCCTGAAAACTATGAAAAGTTTAAAAAGACTACTGCCCGATTAAATGAATTAGAACTATATGAAGCCTATAAGAACTCTTTGCTTCTCGTTATGGAACTTCGAAATCTCGTTGAAAAACAGTATCAATTAACCATTAGCGATGATTTTTTAGGCAAACTGTTTAATTATATGAACGAATAA"}, "NCBI_taxonomy": {"NCBI_taxonomy_name": "Enterococcus faecalis", "NCBI_taxonomy_id": "1351", "NCBI_taxonomy_cvterm_id": "35918"}, "protein_sequence": {"accession": "APU52409.1", "sequence": "MLKQKELMARVKELVQSDERISACMMYGSFTKGEGDQYSDIEYYVFLKDDTISTFDSAKWLNEVASYTLLYQNEYGTEVVIFENLIRGEFHFLSENEMNIIPSFKESGYIPDTKAMFIYDETGQLELYLSELEGPGPNRLTEENVNFLLNNFSNLWLMGINVLKRGENARSLELLSQLQKNILQLIRIAEENADNWFNMTKNLEKEISPENYEKFKKTTARLNELELYEAYKNSLLLVMELRNLVEKQYQLTISDDFLGKLFNYMNE"}}}}}, "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1346": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1347": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1245": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "379": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "378": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "371": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "370": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "373": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "372": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "375": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "374": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "377": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "376": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "393": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "392": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "391": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "390": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "397": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "396": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "395": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "394": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "399": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "398": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1247": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2303": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2306": {"$update": {"model_description": "This model detects protein overexpression based on the presence of mutations.The detection of the protein without an associated mutation indicates that the protein is likely to be expressed at low or basal levels. The detection of the protein with the mutation indicates that the protein is likely overexpressed. This model reflects how certain proteins are functional with and without mutations. For example, efflux pump subunits and regulators are functional with mutations and without mutations. Without mutations, efflux pump subunits and regulators are usually expressed at a low level. When an efflux pump regulator has a mutation, it can cause the overexpression of the efflux pump it is responsible for regulating, leading to resistance to specific drugs. Protein overexpression models have two parameters: a curated BLASTP cutoff, and a curated set of mutations (single resistance variants, frameshift mutations, indels, etc.) shown clinically to confer resistance. This model type is a combination of the protein homolog and protein variant model. A detected hit can be categorized as Perfect, Strict, or Loose with no mutation(s) or as Strict or Loose with mutation(s).", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "375", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "245": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "244": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "247": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "246": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "241": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "240": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "243": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "242": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "249": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "248": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2274": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2277": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2279": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2278": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2154": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "179": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "178": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "177": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "176": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "175": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "174": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "173": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "172": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "800", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "171": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "170": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2051": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2050": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2053": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2052": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2055": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2054": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2057": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2056": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2059": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2058": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1500": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1501": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1506": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1507": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1504": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1977": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1976": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2697": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2695": {"$update": {"model_param": {"$update": {"41141": {"$update": {"param_description": "This detection model parameter describes efflux pump components that are to be detected together (e.g., efflux pump subunits and regulators) using sequential model IDs, separated by commas. For example: 2685,440,1925,1305."}}}}}}, "2694": {"$update": {"model_name": "MexCD\u2013OprJ with type A NfxB mutation", "model_param": {"$update": {"41141": {"$update": {"param_description": "This detection model parameter describes efflux pump components that are to be detected together (e.g., efflux pump subunits and regulators) using sequential model IDs, separated by commas. For example: 2685,440,1925,1305."}}}}}}, "2693": {"$update": {"ARO_description": "Type B NfxB mutants are more resistant to tetracycline and chloramphenicol, as well as ofloxacin, erythromycin, and the new zwitterionic cephems, than was PAO1, and they are four to eight times more susceptible to carbenicillin, sulbenicillin, imipenem, panipenem, biapenem, moxalactam, aztreonam, gentamicin, and kanamycin than PAO1. The mutation at the 46th amino acid position is sufficient for overproduction of OprJ and the multidrug resistance. nfxB corresponds to 2 loci in Pseudomonas aeruginosa PAO1 (gene name: esrC/nfxB) and 2 loci in Pseudomonas aeruginosa LESB58 (gene name: nfxB).", "model_description": "This model detects protein overexpression based on the presence of mutations.The detection of the protein without an associated mutation indicates that the protein is likely to be expressed at low or basal levels. The detection of the protein with the mutation indicates that the protein is likely overexpressed. This model reflects how certain proteins are functional with and without mutations. For example, efflux pump subunits and regulators are functional with mutations and without mutations. Without mutations, efflux pump subunits and regulators are usually expressed at a low level. When an efflux pump regulator has a mutation, it can cause the overexpression of the efflux pump it is responsible for regulating, leading to resistance to specific drugs. Protein overexpression models have two parameters: a curated BLASTP cutoff, and a curated set of mutations (single resistance variants, frameshift mutations, indels, etc.) shown clinically to confer resistance. This model type is a combination of the protein homolog and protein variant model. A detected hit can be categorized as Perfect, Strict, or Loose with no mutation(s) or as Strict or Loose with mutation(s).", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1975": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2691": {"$update": {"ARO_description": "Type A NfxB mutants are four to eight times more resistant to ofloxacin, erythromycin, and new zwitterionic cephems, i.e., cefpirome, cefclidin, cefozopran, and cefoselis, than the parent strain, PAO1. nfxB corresponds to 2 loci in Pseudomonas aeruginosa PAO1 (gene name: esrC/nfxB) and 2 loci in Pseudomonas aeruginosa LESB58 (gene name: nfxB).", "model_description": "This model detects protein overexpression based on the presence of mutations.The detection of the protein without an associated mutation indicates that the protein is likely to be expressed at low or basal levels. The detection of the protein with the mutation indicates that the protein is likely overexpressed. This model reflects how certain proteins are functional with and without mutations. For example, efflux pump subunits and regulators are functional with mutations and without mutations. Without mutations, efflux pump subunits and regulators are usually expressed at a low level. When an efflux pump regulator has a mutation, it can cause the overexpression of the efflux pump it is responsible for regulating, leading to resistance to specific drugs. Protein overexpression models have two parameters: a curated BLASTP cutoff, and a curated set of mutations (single resistance variants, frameshift mutations, indels, etc.) shown clinically to confer resistance. This model type is a combination of the protein homolog and protein variant model. A detected hit can be categorized as Perfect, Strict, or Loose with no mutation(s) or as Strict or Loose with mutation(s).", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1974": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1973": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1972": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1971": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1970": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1968": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1969": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1618": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1619": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1616": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1617": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1614": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1615": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1960": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1613": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1610": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1611": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1768": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1769": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1762": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1763": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1760": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1761": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1766": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1767": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1764": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1765": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1142": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1143": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1140": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1141": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1146": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1147": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1144": {"$update": {"ARO_description": "VgbB inactivates streptogramin B-type antibiotics by linearizing the lactone ring on the ester bond through an elimination mechanism, thus conferring resistance.", "model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "ARO_name": "vgbB", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1145": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1148": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1149": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "690": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "692": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "693": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1544": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "691": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "696": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "697": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "694": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "695": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "698": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "699": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1548": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1549": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "542": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "543": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "540": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "900", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "541": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "546": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "547": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "544": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "545": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "548": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "549": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1782": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1783": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1784": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1785": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1786": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "ARO_name": "mexY", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "1800", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1787": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "414": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "415": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "416": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "417": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "410": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "411": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "412": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "413": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1384": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1385": {"$update": {"ARO_description": "MdsB is the inner membrane transporter of the multidrug and metal efflux complex MdsABC. mdsB corresponds to 1 locus in Pseudomonas aeruginosa PAO1 (gene name: mexQ) and 2 loci in Pseudomonas aeruginosa LESB58.", "model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1386": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1387": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1380": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "419": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1382": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1383": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "368": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "369": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "366": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}, "$delete": ["40334"], "$insert": {"41343": {"param_value": {"8026": "-nt93:5"}, "param_type_id": "41343", "param_type": "deletion mutation from nucleotide sequence", "param_description": "A subtype of the deletion mutation detection model parameter. This parameter is used when a set of deletion mutations is reported in a nucleotide sequence format. Such mutations may be of variable length - possibly causing a frameshift, but not premature termination of functional knockout. Mutation parameters of this type are reported in the CARD with the notation: [-]nt[position]:[nucleotides]."}}}}}, "367": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "364": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "365": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "362": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "363": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "360": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "361": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "380": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "381": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "382": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "383": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "384": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "385": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "386": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "387": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "725", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "388": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "389": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2191": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "258": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2193": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "600", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2194": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "1900", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2195": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "850", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2196": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2198": {"$update": {"ARO_description": "Point mutation in Pseudomonas aeruginosa parE resulting in sensitivity to fluoroquinolones (ciprofloxacin). In combination with a gyrase mutation (gyrA or gyrB), it confers a high level of resistance to ciprofloxacin.", "model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "253": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "250": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "ARO_name": "Rhodococcus fascians cmr", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}, "model_name": "Rhodococcus fascians cmr"}}, "251": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "256": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "257": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "254": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "255": {"$update": {"ARO_description": "A novel bleomycin resistance protein encoded by a metallo-beta-lactamase-associated ble gene. Expression of BRP(MBL) confers resistance to bleomycin and bleomycin-like antibiotics in Enterobacteriaceae and Acinetobacter, where it is co-expressed with an MBL and controlled by the same promoter region.", "model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "ARO_name": "BRP(MBL)", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2200": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2201": {"$update": {"model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2203": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2204": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2205": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2206": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "1900", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2207": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "650", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2208": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "1900", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2428": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2429": {"$update": {"ARO_description": "farB is the cytoplasmic transporter protein that is part of the farAB efflux pump. farB corresponds to 3 loci in Pseudomonas aeruginosa PAO1 and 3 loci in Pseudomonas aeruginosa LESB58.", "model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2400": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2421": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2422": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2423": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "1000", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2424": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_name": "yojI", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1849": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"40394": {"$update": {"param_type": "nonsense mutation", "param_description": "A nucleotide substitution resulting in a change from an amino acid codon to a STOP codon. Nonsense mutations truncate protein translation prematurely, resulting in a defective or completely inactive protein. In CARD, nonsense mutations may be attached to models using the notation: [wild type amino acid][position][STOP] (e.g. Q42STOP). This parameter is not currently used in detection algorithms."}}, "blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}, "$delete": ["40334"], "$insert": {"41345": {"param_value": {"8036": "+nt397:C"}, "param_type_id": "41345", "param_type": "insertion mutation from nucleotide sequence", "param_description": "A subtype of the insertion mutation detection model parameter. This parameter is used when a set of insertion mutations is reported in a nucleotide sequence format. Such mutations may be of variable length - possibly causing a frameshift, but not causing premature termination or a functional knockout. Mutation parameters of this type are reported in CARD with the notation: [+]nt[position]:[nucleotides]."}, "41343": {"param_value": {"8035": "-nt310:G", "8034": "-nt26:C", "8031": "-nt477:G", "8030": "-nt586:G", "8033": "-nt23:A", "8032": "-nt400:A", "8028": "-nt673:GT", "8029": "-nt653:T", "8027": "-nt758:G"}, "param_type_id": "41343", "param_type": "deletion mutation from nucleotide sequence", "param_description": "A subtype of the deletion mutation detection model parameter. This parameter is used when a set of deletion mutations is reported in a nucleotide sequence format. Such mutations may be of variable length - possibly causing a frameshift, but not premature termination of functional knockout. Mutation parameters of this type are reported in the CARD with the notation: [-]nt[position]:[nucleotides]."}}}}}, "2426": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2427": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2432": {"$update": {"model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1848": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "168": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "169": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "164": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "165": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "166": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "167": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "160": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "161": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "162": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "163": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2518": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2519": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1980": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2517": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1841": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2734": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1840": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2731": {"$update": {"model_param": {"$update": {"41141": {"$update": {"param_description": "This detection model parameter describes efflux pump components that are to be detected together (e.g., efflux pump subunits and regulators) using sequential model IDs, separated by commas. For example: 2685,440,1925,1305."}}}}}}, "2732": {"$update": {"model_param": {"$update": {"41141": {"$update": {"param_description": "This detection model parameter describes efflux pump components that are to be detected together (e.g., efflux pump subunits and regulators) using sequential model IDs, separated by commas. For example: 2685,440,1925,1305."}}}}}}, "2733": {"$update": {"model_param": {"$update": {"41141": {"$update": {"param_description": "This detection model parameter describes efflux pump components that are to be detected together (e.g., efflux pump subunits and regulators) using sequential model IDs, separated by commas. For example: 2685,440,1925,1305."}}}}}}, "678": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "679": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1814": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1815": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1816": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1817": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1810": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1811": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1812": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1813": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1818": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1819": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "670": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "671": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1609": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1608": {"$update": {"model_description": "This model detects protein overexpression based on the presence of mutations.The detection of the protein without an associated mutation indicates that the protein is likely to be expressed at low or basal levels. The detection of the protein with the mutation indicates that the protein is likely overexpressed. This model reflects how certain proteins are functional with and without mutations. For example, efflux pump subunits and regulators are functional with mutations and without mutations. Without mutations, efflux pump subunits and regulators are usually expressed at a low level. When an efflux pump regulator has a mutation, it can cause the overexpression of the efflux pump it is responsible for regulating, leading to resistance to specific drugs. Protein overexpression models have two parameters: a curated BLASTP cutoff, and a curated set of mutations (single resistance variants, frameshift mutations, indels, etc.) shown clinically to confer resistance. This model type is a combination of the protein homolog and protein variant model. A detected hit can be categorized as Perfect, Strict, or Loose with no mutation(s) or as Strict or Loose with mutation(s).", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "500", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1979": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1978": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1601": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1600": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1603": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1602": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1605": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1604": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1607": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1606": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "809": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "808": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "803": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "802": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "801": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "800": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "807": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "806": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "805": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "600", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "804": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1775": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1774": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1777": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1776": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1771": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1770": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1773": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1772": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1779": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1778": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1159": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1158": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1155": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1154": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1157": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1156": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1151": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1150": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1153": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1152": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1552": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1555": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1554": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1551": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1550": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1553": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1101": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "59": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "58": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1557": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1556": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "55": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "54": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "57": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "56": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "51": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "50": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "53": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "52": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "537": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "536": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "535": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "534": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "533": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "532": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "531": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "530": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "539": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "538": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1558": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "429": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "428": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1399": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1398": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1397": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "420": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "423": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "422": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "425": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1392": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "427": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1390": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2183": {"$update": {"model_param": {"$update": {"40297": {"$update": {"param_description": "The gene order model parameter describes the relative order (5'->3') of a set of genes or other genetic elements on a chromosome, a plasmid or within an operon. Antibiotic resistance is only conferred when the detected set of genes appears in the indicated order; otherwise, no resistance phenotype is produced. This parameter is part of the gene cluster meta-model, and may be attached to detection models with the following notation: [[cvterm_id 1],[cvterm_id 2],...,[cvterm_id n]], where the cvterm_id denotes a gene-associated AMR term and an attached model id. This parameter currently (August 2017) lacks an algorithm for detection."}}}}}}, "2182": {"$update": {"model_param": {"$update": {"40297": {"$update": {"param_description": "The gene order model parameter describes the relative order (5'->3') of a set of genes or other genetic elements on a chromosome, a plasmid or within an operon. Antibiotic resistance is only conferred when the detected set of genes appears in the indicated order; otherwise, no resistance phenotype is produced. This parameter is part of the gene cluster meta-model, and may be attached to detection models with the following notation: [[cvterm_id 1],[cvterm_id 2],...,[cvterm_id n]], where the cvterm_id denotes a gene-associated AMR term and an attached model id. This parameter currently (August 2017) lacks an algorithm for detection."}}}}}}, "2181": {"$update": {"model_param": {"$update": {"40297": {"$update": {"param_description": "The gene order model parameter describes the relative order (5'->3') of a set of genes or other genetic elements on a chromosome, a plasmid or within an operon. Antibiotic resistance is only conferred when the detected set of genes appears in the indicated order; otherwise, no resistance phenotype is produced. This parameter is part of the gene cluster meta-model, and may be attached to detection models with the following notation: [[cvterm_id 1],[cvterm_id 2],...,[cvterm_id n]], where the cvterm_id denotes a gene-associated AMR term and an attached model id. This parameter currently (August 2017) lacks an algorithm for detection."}}}}}}, "2180": {"$update": {"model_param": {"$update": {"40297": {"$update": {"param_description": "The gene order model parameter describes the relative order (5'->3') of a set of genes or other genetic elements on a chromosome, a plasmid or within an operon. Antibiotic resistance is only conferred when the detected set of genes appears in the indicated order; otherwise, no resistance phenotype is produced. This parameter is part of the gene cluster meta-model, and may be attached to detection models with the following notation: [[cvterm_id 1],[cvterm_id 2],...,[cvterm_id n]], where the cvterm_id denotes a gene-associated AMR term and an attached model id. This parameter currently (August 2017) lacks an algorithm for detection."}}}}}}, "2186": {"$update": {"model_param": {"$update": {"40297": {"$update": {"param_description": "The gene order model parameter describes the relative order (5'->3') of a set of genes or other genetic elements on a chromosome, a plasmid or within an operon. Antibiotic resistance is only conferred when the detected set of genes appears in the indicated order; otherwise, no resistance phenotype is produced. This parameter is part of the gene cluster meta-model, and may be attached to detection models with the following notation: [[cvterm_id 1],[cvterm_id 2],...,[cvterm_id n]], where the cvterm_id denotes a gene-associated AMR term and an attached model id. This parameter currently (August 2017) lacks an algorithm for detection."}}}}}}, "229": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "228": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "227": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "226": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "225": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "224": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "223": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "222": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "221": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "220": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2213": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "850", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2212": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "1900", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2211": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "650", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2217": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "1900", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2216": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "650", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2215": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2219": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "151": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "150": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "350", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "153": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "152": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "155": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "154": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "157": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "156": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "159": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "158": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2431": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2430": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2436": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}, "$delete": ["40334"], "$insert": {"41345": {"param_value": {"8151": "+nt41:GAGCA"}, "param_type_id": "41345", "param_type": "insertion mutation from nucleotide sequence", "param_description": "A subtype of the insertion mutation detection model parameter. This parameter is used when a set of insertion mutations is reported in a nucleotide sequence format. Such mutations may be of variable length - possibly causing a frameshift, but not causing premature termination or a functional knockout. Mutation parameters of this type are reported in CARD with the notation: [+]nt[position]:[nucleotides]."}, "41342": {"param_value": {"4529": "-DV244-245"}, "param_type_id": "41342", "param_type": "deletion mutation from peptide sequence", "param_description": "A subtype of the deletion mutation detection model parameter. This parameter is used when a set of deletion mutations is reported in a peptide sequence format. These are specific to codon deletions, where a multiple of 3 nucleotides are deleted. Mutations of this type are reported in the CARD with the notation: [-][AAs][position range]."}}}}}, "2435": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2434": {"$update": {"model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2724": {"$update": {"model_param": {"$update": {"41141": {"$update": {"param_description": "This detection model parameter describes efflux pump components that are to be detected together (e.g., efflux pump subunits and regulators) using sequential model IDs, separated by commas. For example: 2685,440,1925,1305."}}}}}}, "2720": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "1900", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2729": {"$update": {"model_param": {"$update": {"41141": {"$update": {"param_description": "This detection model parameter describes efflux pump components that are to be detected together (e.g., efflux pump subunits and regulators) using sequential model IDs, separated by commas. For example: 2685,440,1925,1305."}}}}}}, "1807": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1806": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1805": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1804": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1803": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1802": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1801": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1800": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1809": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1808": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1524": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1948": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1949": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1525": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1942": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1943": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1940": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1941": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1946": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1947": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1944": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1945": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "818": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "819": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1527": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "810": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "811": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "812": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "813": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "814": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "815": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "816": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "817": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1991": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1522": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1990": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1523": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "1800", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1993": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1490": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "421": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1492": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1493": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1494": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1495": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1496": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1497": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1498": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1499": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1395": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1994": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1700": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1701": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1702": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1703": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1704": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1705": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1706": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1707": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1708": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1709": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1996": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "424": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1391": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "426": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1128": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1129": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1120": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1121": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1122": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1123": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1124": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1125": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1126": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1127": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "524": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "525": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "526": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "527": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1018": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "521": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "522": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "523": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1014": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1015": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1016": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1017": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "528": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "529": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1012": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1013": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1234": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1235": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1236": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1237": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}, "40394": {"$update": {"param_type": "nonsense mutation", "param_description": "A nucleotide substitution resulting in a change from an amino acid codon to a STOP codon. Nonsense mutations truncate protein translation prematurely, resulting in a defective or completely inactive protein. In CARD, nonsense mutations may be attached to models using the notation: [wild type amino acid][position][STOP] (e.g. Q42STOP). This parameter is not currently used in detection algorithms."}}}, "$delete": ["40334"], "$insert": {"41344": {"param_value": {"4017": "+F514", "3999": "+R514"}, "param_type_id": "41344", "param_type": "insertion mutation from peptide sequence", "param_description": "A subtype of the insertion mutation detection model parameter. This parameter is used when a set of insertion mutations is reported in a peptide sequence format. These are specific to codon insertions, where a multiple of three nucleotides are inserted. This does not cause a frameshift mutation. Mutation parameters of this type are reported in CARD with the notation: [+][AAs][position range]."}, "41342": {"param_value": {"8084": "-P520", "8085": "-Q517", "8086": "-N519", "8083": "-K527", "4006": "-N516", "4026": "-DQ516-517", "4027": "-QN517-518"}, "param_type_id": "41342", "param_type": "deletion mutation from peptide sequence", "param_description": "A subtype of the deletion mutation detection model parameter. This parameter is used when a set of deletion mutations is reported in a peptide sequence format. These are specific to codon deletions, where a multiple of 3 nucleotides are deleted. Mutations of this type are reported in the CARD with the notation: [-][AAs][position range]."}}}}}, "1230": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1231": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1232": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1233": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1238": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1239": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "_version": "1.2.0", "438": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "439": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "436": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "437": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "434": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "435": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "432": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "433": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "430": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "431": {"$update": {"model_description": "This model detects protein overexpression based on the presence of mutations.The detection of the protein without an associated mutation indicates that the protein is likely to be expressed at low or basal levels. The detection of the protein with the mutation indicates that the protein is likely overexpressed. This model reflects how certain proteins are functional with and without mutations. For example, efflux pump subunits and regulators are functional with mutations and without mutations. Without mutations, efflux pump subunits and regulators are usually expressed at a low level. When an efflux pump regulator has a mutation, it can cause the overexpression of the efflux pump it is responsible for regulating, leading to resistance to specific drugs. Protein overexpression models have two parameters: a curated BLASTP cutoff, and a curated set of mutations (single resistance variants, frameshift mutations, indels, etc.) shown clinically to confer resistance. This model type is a combination of the protein homolog and protein variant model. A detected hit can be categorized as Perfect, Strict, or Loose with no mutation(s) or as Strict or Loose with mutation(s).", "model_param": {"$update": {"40394": {"$update": {"param_type": "nonsense mutation", "param_description": "A nucleotide substitution resulting in a change from an amino acid codon to a STOP codon. Nonsense mutations truncate protein translation prematurely, resulting in a defective or completely inactive protein. In CARD, nonsense mutations may be attached to models using the notation: [wild type amino acid][position][STOP] (e.g. Q42STOP). This parameter is not currently used in detection algorithms."}}, "blastp_bit_score": {"$update": {"param_value": "210", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1967": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1961": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "238": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "239": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "234": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "235": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "236": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "237": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "230": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "231": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "232": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "233": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2462": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2228": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2229": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2227": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2224": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2222": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2223": {"$update": {"model_description": "This model detects protein overexpression based on the presence of mutations.The detection of the protein without an associated mutation indicates that the protein is likely to be expressed at low or basal levels. The detection of the protein with the mutation indicates that the protein is likely overexpressed. This model reflects how certain proteins are functional with and without mutations. For example, efflux pump subunits and regulators are functional with mutations and without mutations. Without mutations, efflux pump subunits and regulators are usually expressed at a low level. When an efflux pump regulator has a mutation, it can cause the overexpression of the efflux pump it is responsible for regulating, leading to resistance to specific drugs. Protein overexpression models have two parameters: a curated BLASTP cutoff, and a curated set of mutations (single resistance variants, frameshift mutations, indels, etc.) shown clinically to confer resistance. This model type is a combination of the protein homolog and protein variant model. A detected hit can be categorized as Perfect, Strict, or Loose with no mutation(s) or as Strict or Loose with mutation(s).", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2221": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "146": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "147": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "144": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "145": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "142": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "143": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "140": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "141": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "148": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "149": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2088": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "2083": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2080": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "2081": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "800", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2086": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "2087": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2084": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "2085": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "2712": {"$update": {"model_param": {"$update": {"41141": {"$update": {"param_description": "This detection model parameter describes efflux pump components that are to be detected together (e.g., efflux pump subunits and regulators) using sequential model IDs, separated by commas. For example: 2685,440,1925,1305."}}}}}}, "2713": {"$update": {"model_param": {"$update": {"41141": {"$update": {"param_description": "This detection model parameter describes efflux pump components that are to be detected together (e.g., efflux pump subunits and regulators) using sequential model IDs, separated by commas. For example: 2685,440,1925,1305."}}}}}}, "2711": {"$update": {"model_param": {"$update": {"41141": {"$update": {"param_description": "This detection model parameter describes efflux pump components that are to be detected together (e.g., efflux pump subunits and regulators) using sequential model IDs, separated by commas. For example: 2685,440,1925,1305."}}}}}}, "2716": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "850", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2717": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2718": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "1900", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1832": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1833": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1830": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1831": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1836": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1837": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1834": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1835": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1838": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1839": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2406": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2155": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "2156": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2157": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "2402": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2403": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2152": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "2401": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "933": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "932": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "931": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "937": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "936": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "935": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2409": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1955": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1954": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1957": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1956": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1951": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1950": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1953": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1952": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1959": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1958": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "829": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "828": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "825": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "824": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "827": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "826": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "900", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "821": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "820": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "1800", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "823": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "822": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1483": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1482": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1481": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1480": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1487": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1486": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1485": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1484": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1489": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1488": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "797": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2411": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "795": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "794": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "793": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "792": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "791": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "929": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1719": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1718": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "799": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "798": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1270": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2412": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "613": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1272": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1139": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1138": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1133": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "616": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1131": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1130": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1137": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1136": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1135": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "617": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2792": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "ARO_category": {"$insert": {"36454": {"category_aro_name": "determinant of macrolide resistance", "category_aro_cvterm_id": "36454", "category_aro_accession": "3000315", "category_aro_description": "Enzymes, other proteins or other gene products shown clinically to confer resistance to macrolide antibiotics."}}}, "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "1276": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2793": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "ARO_category": {"$insert": {"36454": {"category_aro_name": "determinant of macrolide resistance", "category_aro_cvterm_id": "36454", "category_aro_accession": "3000315", "category_aro_description": "Enzymes, other proteins or other gene products shown clinically to confer resistance to macrolide antibiotics."}}}, "ARO_name": "Chlamydomonas reinhardtii 23S rRNA with mutation conferring resistance to erythromycin", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "476": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "1750", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1277": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2790": {"$update": {"ARO_description": "Omp1 is an outer membrane porin that confers resistance by absence in S. marcescens.  Knockout, deletion or other inhibition of the omp1 gene confers resistance to certain beta-lactamase antibiotics - including Cefoxitin, Ceftriaxone, Cefotaxime, and Moxalactam - as well as Ciprofloxacin, Tetracycline, and Chloramphenicol, by preventing passage of the antibiotic into the cell."}, "$insert": {"model_param": {"blastp_bit_score": {"param_value": "700", "param_type_id": "40725", "param_type": "BLASTP bit-score", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}, "2791": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2796": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2797": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "519": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "518": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "926": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1009": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1008": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1007": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1006": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "513": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1004": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "515": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1002": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1001": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1000": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "623": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "622": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1225": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "620": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1223": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "626": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "625": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1220": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "629": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "628": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1229": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1228": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1535": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1561": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1286": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "11": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "10": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "13": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "12": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "15": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "14": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "17": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "16": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "19": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "18": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "928": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1534": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "201": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "200": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "203": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "202": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "205": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "204": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "207": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "206": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "209": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "208": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1573": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1572": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1571": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1570": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2231": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2230": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2233": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2232": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2235": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2234": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1576": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1575": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1574": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2185": {"$update": {"model_param": {"$update": {"40297": {"$update": {"param_description": "The gene order model parameter describes the relative order (5'->3') of a set of genes or other genetic elements on a chromosome, a plasmid or within an operon. Antibiotic resistance is only conferred when the detected set of genes appears in the indicated order; otherwise, no resistance phenotype is produced. This parameter is part of the gene cluster meta-model, and may be attached to detection models with the following notation: [[cvterm_id 1],[cvterm_id 2],...,[cvterm_id n]], where the cvterm_id denotes a gene-associated AMR term and an attached model id. This parameter currently (August 2017) lacks an algorithm for detection."}}}}}}, "2184": {"$update": {"model_param": {"$update": {"40297": {"$update": {"param_description": "The gene order model parameter describes the relative order (5'->3') of a set of genes or other genetic elements on a chromosome, a plasmid or within an operon. Antibiotic resistance is only conferred when the detected set of genes appears in the indicated order; otherwise, no resistance phenotype is produced. This parameter is part of the gene cluster meta-model, and may be attached to detection models with the following notation: [[cvterm_id 1],[cvterm_id 2],...,[cvterm_id n]], where the cvterm_id denotes a gene-associated AMR term and an attached model id. This parameter currently (August 2017) lacks an algorithm for detection."}}}}}}, "2097": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "2096": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "2091": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "2090": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "2093": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "2092": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"40494": {"$update": {"param_type": "frameshift mutation", "param_description": "A frameshift is a type of genetic mutation caused by a nucleotide insertion or deletion \u2260 3 bases. This changes the grouping of codons and thus the reading frame during translation, resulting in a incomplete or inactive protein product. Many frameshift mutations generate downstream STOP codons, resulting in premature peptide translation termination. Frameshifts may also confer antibiotic resistance through partial or total protein loss-of-function. Frameshift mutations are included with relevant models when applicable, with the following notation: [wild-type AA][position]fs;[[wild-type AA][position]STOP], where AA is an amino acid. If the premature STOP codon position is unknown or does not exist, [wild-type AA][position]fs is sufficient. This parameter is currently not included in detection algorithms."}}, "blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}, "$delete": ["40334"], "$insert": {"41343": {"param_value": {"8024": "-nt139:1"}, "param_type_id": "41343", "param_type": "deletion mutation from nucleotide sequence", "param_description": "A subtype of the deletion mutation detection model parameter. This parameter is used when a set of deletion mutations is reported in a nucleotide sequence format. Such mutations may be of variable length - possibly causing a frameshift, but not premature termination of functional knockout. Mutation parameters of this type are reported in the CARD with the notation: [-]nt[position]:[nucleotides]."}}}}}, "2099": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "2098": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "2525": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2524": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2527": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2526": {"$update": {"ARO_description": "VgbC is a streptogramin B-type lyase found in Paenibacillus sp. LC231, a strain of Paenibacillus isolated from Lechuguilla Cave, NM, USA. Confers resistance to streptogramin B-type antibiotics by linearization of the lactone ring on an ester bond, resulting in antibiotic inactivation. Described by Pawlowski et al. 2016.", "model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "ARO_name": "vgbC", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2521": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2520": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2523": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2522": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2529": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2528": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2705": {"$update": {"ARO_name": "MexEF-OprN with MexS mutations conferring resistance to chloramphenicol, ciprofloxacin, and trimethoprim", "model_param": {"$update": {"41141": {"$update": {"param_description": "This detection model parameter describes efflux pump components that are to be detected together (e.g., efflux pump subunits and regulators) using sequential model IDs, separated by commas. For example: 2685,440,1925,1305."}}}}, "model_name": "MexEF-OprN with MexS mutations conferring resistance to chloramphenicol, ciprofloxacin, and trimethoprim"}}, "2704": {"$update": {"ARO_name": "MexEF-OprN with MexT mutation conferring resistance to chloramphenicol, ciprofloxacin, and trimethoprim", "model_param": {"$update": {"41141": {"$update": {"param_description": "This detection model parameter describes efflux pump components that are to be detected together (e.g., efflux pump subunits and regulators) using sequential model IDs, separated by commas. For example: 2685,440,1925,1305."}}}}, "model_name": "MexEF-OprN with MexT mutation conferring resistance to chloramphenicol, ciprofloxacin, and trimethoprim"}}, "2707": {"$update": {"model_param": {"$update": {"41141": {"$update": {"param_description": "This detection model parameter describes efflux pump components that are to be detected together (e.g., efflux pump subunits and regulators) using sequential model IDs, separated by commas. For example: 2685,440,1925,1305."}}}}}}, "2706": {"$update": {"model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1829": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1828": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1825": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1824": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1827": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1826": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1821": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1820": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1823": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1822": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2147": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2146": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "2145": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "2144": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2143": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "2142": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "2141": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "2140": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "920": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "921": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "922": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "923": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "924": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "925": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2149": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "2148": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1920": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1921": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1922": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1923": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1924": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1925": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "1950", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1926": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1927": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1928": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "300", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1929": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "475": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "832": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "833": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "830": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "831": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "836": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "837": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "834": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "835": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "838": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "839": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "3": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1986": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1987": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1532": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "784": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "785": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "786": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "787": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "780": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "781": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "782": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1729": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1726": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1727": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1724": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1725": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "788": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "789": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1720": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1721": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "252": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "60": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "61": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "62": {"$update": {"ARO_description": "CMY-42 is a plasmid-borne AmpC cephalosporinase gene found in Escherichia coli", "model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "63": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "64": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "65": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "66": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "67": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "68": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "69": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1371": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1588": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1589": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "406": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1582": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1583": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1580": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1581": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1586": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1373": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1584": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1585": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2433": {"$update": {"ARO_description": "lfrA is involved in the active efflux of quinolones and is found in Mycobacterium smegmatis. lfrA corresponds to 2 loci in Pseudomonas aeruginosa PAO1 and 2 loci in Pseudomonas aeruginosa LESB58.", "model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1038": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "404": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "508": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "509": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1032": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "507": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "504": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1031": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "502": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "503": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1034": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "402": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1212": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "631": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "632": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1211": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1216": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1217": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "636": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "637": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "638": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "639": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1218": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "927": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1394": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2416": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "465": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1728": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "783": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1106": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1455": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1104": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "1800", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1105": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1450": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1103": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1452": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1453": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1458": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1459": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1108": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1109": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1722": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1723": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1577": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "466": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2136": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "2137": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2794": {"$update": {"ARO_description": "Point mutations in Helicobacter pylori shown to confer resistance to clarithromycin, a macrolide antibiotic", "model_type": "rRNA gene variant model", "ARO_category": {"$insert": {"36454": {"category_aro_name": "determinant of macrolide resistance", "category_aro_cvterm_id": "36454", "category_aro_accession": "3000315", "category_aro_description": "Enzymes, other proteins or other gene products shown clinically to confer resistance to macrolide antibiotics."}}}, "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}, "snp": {"$update": {"param_value": {"7908": "A2147G", "7909": "A2147C", "7905": "A2146C", "7906": "A2146G", "7907": "A2146U"}, "clinical": {"7908": "A2147G", "7909": "A2147C", "7905": "A2146C", "7906": "A2146G", "7907": "A2146U"}}}}}, "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected."}}, "1393": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "216": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "217": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "214": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "215": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "213": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "210": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "211": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1530": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "218": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "219": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2138": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2139": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "462": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}, "$delete": ["40334"], "$insert": {"41344": {"param_value": {"3863": "+G76", "3864": "+E77"}, "param_type_id": "41344", "param_type": "insertion mutation from peptide sequence", "param_description": "A subtype of the insertion mutation detection model parameter. This parameter is used when a set of insertion mutations is reported in a peptide sequence format. These are specific to codon insertions, where a multiple of three nucleotides are inserted. This does not cause a frameshift mutation. Mutation parameters of this type are reported in CARD with the notation: [+][AAs][position range]."}}}}}, "939": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "4": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "938": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2550": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2551": {"$update": {"model_param": {"$update": {"40297": {"$update": {"param_description": "The gene order model parameter describes the relative order (5'->3') of a set of genes or other genetic elements on a chromosome, a plasmid or within an operon. Antibiotic resistance is only conferred when the detected set of genes appears in the indicated order; otherwise, no resistance phenotype is produced. This parameter is part of the gene cluster meta-model, and may be attached to detection models with the following notation: [[cvterm_id 1],[cvterm_id 2],...,[cvterm_id n]], where the cvterm_id denotes a gene-associated AMR term and an attached model id. This parameter currently (August 2017) lacks an algorithm for detection."}}}}}}, "2396": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2397": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2395": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2398": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2399": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2778": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2779": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2770": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2771": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2773": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2774": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2775": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2776": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1858": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1859": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1850": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1851": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1852": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1853": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1854": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1855": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1856": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1857": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "919": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "918": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "915": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "914": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "917": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "916": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "911": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "910": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "913": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "912": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1933": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1932": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1931": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1930": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1937": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1936": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1935": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1934": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1939": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1938": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "847": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "846": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "845": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "844": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "843": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "842": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_name": "pmrE", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "841": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "840": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "849": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "848": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1587": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2407": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1739": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1738": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1731": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1730": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1733": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1732": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1735": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1734": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1737": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1736": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1039": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "796": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "753": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "752": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "751": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "750": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "757": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "756": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "755": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "754": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "759": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "758": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1595": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "506": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1597": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1596": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1591": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1590": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1593": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1592": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1599": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1030": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1025": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1024": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1027": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1026": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1021": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1020": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1023": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1022": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1036": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1029": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1028": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1037": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "500": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "501": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "605": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "604": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "607": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "606": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "601": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "600": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "603": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "600", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "602": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1205": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1204": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1207": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1206": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "609": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "600", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1200": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1203": {"$update": {"ARO_description": "Point mutations in the Mycobacterium tuberculosis ndh gene shown clinically to confer resistance to isoniazid.", "model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "ARO_name": "Mycobacterium tuberculosis ndh with mutation conferring resistance to isoniazid", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1202": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "633": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "634": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "635": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1214": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1215": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1111": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1110": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1113": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1112": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1115": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1114": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1117": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1440": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1119": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1118": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "467": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1449": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1448": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1219": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1357": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "460": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1355": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "489": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"40394": {"$update": {"param_type": "nonsense mutation", "param_description": "A nucleotide substitution resulting in a change from an amino acid codon to a STOP codon. Nonsense mutations truncate protein translation prematurely, resulting in a defective or completely inactive protein. In CARD, nonsense mutations may be attached to models using the notation: [wild type amino acid][position][STOP] (e.g. Q42STOP). This parameter is not currently used in detection algorithms."}}, "blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}, "$delete": ["40334"], "$insert": {"41345": {"param_value": {"8037": "+nt350:G", "8038": "+nt351:G"}, "param_type_id": "41345", "param_type": "insertion mutation from nucleotide sequence", "param_description": "A subtype of the insertion mutation detection model parameter. This parameter is used when a set of insertion mutations is reported in a nucleotide sequence format. Such mutations may be of variable length - possibly causing a frameshift, but not causing premature termination or a functional knockout. Mutation parameters of this type are reported in CARD with the notation: [+]nt[position]:[nucleotides]."}, "41343": {"param_value": {"8039": "-nt115:C"}, "param_type_id": "41343", "param_type": "deletion mutation from nucleotide sequence", "param_description": "A subtype of the deletion mutation detection model parameter. This parameter is used when a set of deletion mutations is reported in a nucleotide sequence format. Such mutations may be of variable length - possibly causing a frameshift, but not premature termination of functional knockout. Mutation parameters of this type are reported in the CARD with the notation: [-]nt[position]:[nucleotides]."}}}}}, "488": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "487": {"$update": {"ARO_description": "MacB is an ATP-binding cassette (ABC) transporter that exports macrolides with 14- or 15- membered lactones. It forms an antibiotic efflux complex with MacA and TolC. macB corresponds to 1 locus in Pseudomonas aeruginosa PAO1 and 1 locus in Pseudomonas aeruginosa LESB58.", "model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "486": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "485": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "484": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "483": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "482": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "481": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "480": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "790": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "199": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "198": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "195": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "194": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "197": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "196": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "191": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "190": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "193": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "192": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1454": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1107": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1456": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2383": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1457": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2387": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2386": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1102": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1451": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2769": {"$update": {"model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2768": {"$update": {"model_param": {"$update": {"41141": {"$update": {"param_description": "This detection model parameter describes efflux pump components that are to be detected together (e.g., efflux pump subunits and regulators) using sequential model IDs, separated by commas. For example: 2685,440,1925,1305."}}}}}}, "1100": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2763": {"$update": {"model_param": {"$update": {"41141": {"$update": {"param_description": "This detection model parameter describes efflux pump components that are to be detected together (e.g., efflux pump subunits and regulators) using sequential model IDs, separated by commas. For example: 2685,440,1925,1305."}}}}}}, "2762": {"$update": {"model_param": {"$update": {"41141": {"$update": {"param_description": "This detection model parameter describes efflux pump components that are to be detected together (e.g., efflux pump subunits and regulators) using sequential model IDs, separated by commas. For example: 2685,440,1925,1305."}}}}}}, "2761": {"$update": {"model_param": {"$update": {"41141": {"$update": {"param_description": "This detection model parameter describes efflux pump components that are to be detected together (e.g., efflux pump subunits and regulators) using sequential model IDs, separated by commas. For example: 2685,440,1925,1305."}}}}}}, "2760": {"$update": {"model_param": {"$update": {"41141": {"$update": {"param_description": "This detection model parameter describes efflux pump components that are to be detected together (e.g., efflux pump subunits and regulators) using sequential model IDs, separated by commas. For example: 2685,440,1925,1305."}}}}}}, "2766": {"$update": {"model_param": {"$update": {"41141": {"$update": {"param_description": "This detection model parameter describes efflux pump components that are to be detected together (e.g., efflux pump subunits and regulators) using sequential model IDs, separated by commas. For example: 2685,440,1925,1305."}}}}}}, "2765": {"$update": {"model_param": {"$update": {"41141": {"$update": {"param_description": "This detection model parameter describes efflux pump components that are to be detected together (e.g., efflux pump subunits and regulators) using sequential model IDs, separated by commas. For example: 2685,440,1925,1305."}}}}}}, "2764": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "902": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "903": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "900": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "901": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "906": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "907": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "904": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "905": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1843": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1842": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "908": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "909": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1847": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "800", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1846": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1845": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1844": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2614": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1908": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1909": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1906": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1907": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1904": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1905": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1902": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1903": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "675", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1900": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1901": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "854": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "855": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "856": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "850": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "851": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"40394": {"$update": {"param_type": "nonsense mutation", "param_description": "A nucleotide substitution resulting in a change from an amino acid codon to a STOP codon. Nonsense mutations truncate protein translation prematurely, resulting in a defective or completely inactive protein. In CARD, nonsense mutations may be attached to models using the notation: [wild type amino acid][position][STOP] (e.g. Q42STOP). This parameter is not currently used in detection algorithms."}}, "blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}, "40330": {"$update": {"param_value": {"$delete": ["4306", "4285"], "$insert": {"8095": "V130G,+nt420:GG", "8103": "D53N,+nt349:CACTG"}}}}}, "$delete": ["40334"], "$insert": {"41345": {"param_value": {"8105": "+nt407:C", "8096": "+nt420:G", "8100": "+nt493:C", "8102": "+nt218:CGCATTGCCG", "8116": "+nt287:T", "8109": "+nt532:C", "8108": "+nt392:G", "8104": "+nt221:G", "8098": "+nt193:A", "8099": "+nt52:G", "8129": "+nt403:C", "8128": "+nt388:AGGTCGATG", "8112": "+nt414:G", "8113": "+nt391:GG", "8110": "+nt480:TGAC", "8111": "+nt465:T", "8130": "+nt397:T", "8131": "+nt475:C", "8114": "+nt382:AG", "8115": "+nt368:18"}, "param_type_id": "41345", "param_type": "insertion mutation from nucleotide sequence", "param_description": "A subtype of the insertion mutation detection model parameter. This parameter is used when a set of insertion mutations is reported in a nucleotide sequence format. Such mutations may be of variable length - possibly causing a frameshift, but not causing premature termination or a functional knockout. Mutation parameters of this type are reported in CARD with the notation: [+]nt[position]:[nucleotides]."}, "41343": {"param_value": {"8118": "-nt406:G", "8119": "-nt386:ATGT", "8138": "-nt391:G", "8139": "-nt104:C", "8134": "-nt446:8", "8135": "-nt443:G", "8136": "-nt1:11", "8137": "-nt416:TG", "8117": "-nt452:T", "8132": "-nt70:G", "8133": "-nt518:5", "8097": "-nt301:G", "8107": "-nt512:C", "8106": "-nt195:68", "8101": "-nt386:ATG", "8141": "-nt28:C", "8140": "-nt71:G", "8143": "-nt84:C", "8142": "-nt379:11", "8127": "-nt267:24", "8126": "-nt77:G", "8125": "-nt151:80", "8124": "-nt158:A", "8123": "-nt161:C", "8122": "-nt307:TACAG", "8121": "-nt341:C", "8120": "-nt381:GG"}, "param_type_id": "41343", "param_type": "deletion mutation from nucleotide sequence", "param_description": "A subtype of the deletion mutation detection model parameter. This parameter is used when a set of deletion mutations is reported in a nucleotide sequence format. Such mutations may be of variable length - possibly causing a frameshift, but not premature termination of functional knockout. Mutation parameters of this type are reported in the CARD with the notation: [-]nt[position]:[nucleotides]."}}}}}, "852": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "853": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "858": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "740": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "741": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "742": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "743": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "744": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "745": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "746": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "747": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "748": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "749": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1050": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1051": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1052": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1053": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1055": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1056": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1057": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1058": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1059": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1696": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1697": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1694": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1695": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1692": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1693": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1690": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1691": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1791": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1698": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1699": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1278": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1279": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "618": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "619": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "612": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1271": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "610": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "611": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1274": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1275": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "614": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "615": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1795": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1794": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1491": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1472": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1473": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1470": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1471": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1476": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1477": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1474": {"$update": {"model_description": "This model detects protein overexpression based on the presence of mutations.The detection of the protein without an associated mutation indicates that the protein is likely to be expressed at low or basal levels. The detection of the protein with the mutation indicates that the protein is likely overexpressed. This model reflects how certain proteins are functional with and without mutations. For example, efflux pump subunits and regulators are functional with mutations and without mutations. Without mutations, efflux pump subunits and regulators are usually expressed at a low level. When an efflux pump regulator has a mutation, it can cause the overexpression of the efflux pump it is responsible for regulating, leading to resistance to specific drugs. Protein overexpression models have two parameters: a curated BLASTP cutoff, and a curated set of mutations (single resistance variants, frameshift mutations, indels, etc.) shown clinically to confer resistance. This model type is a combination of the protein homolog and protein variant model. A detected hit can be categorized as Perfect, Strict, or Loose with no mutation(s) or as Strict or Loose with mutation(s).", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1475": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1478": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1479": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1304": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1305": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "850", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1306": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1307": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1300": {"$update": {"ARO_description": "MexF is the multidrug inner membrane transporter of the MexEF-OprN complex. mexF corresponds to 2 loci in Pseudomonas aeruginosa PAO1 (gene name: mexF/mexB) and 4 loci in Pseudomonas aeruginosa LESB58 (gene name: mexD/mexB).", "model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1301": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1302": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1303": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1308": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1309": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "498": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "499": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "494": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "495": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "496": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "497": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "490": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "491": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "492": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "493": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "24": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "25": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "26": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "27": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "20": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "21": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "22": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "23": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "28": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "29": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1241": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "7": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2281": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2282": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2283": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2284": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2375": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2372": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}, "$delete": ["40334"], "$insert": {"41343": {"param_value": {"8144": "-nt602:20"}, "param_type_id": "41343", "param_type": "deletion mutation from nucleotide sequence", "param_description": "A subtype of the deletion mutation detection model parameter. This parameter is used when a set of deletion mutations is reported in a nucleotide sequence format. Such mutations may be of variable length - possibly causing a frameshift, but not premature termination of functional knockout. Mutation parameters of this type are reported in the CARD with the notation: [-]nt[position]:[nucleotides]."}}}}}, "2373": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1087": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1086": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1085": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1876": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1877": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1874": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1875": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1872": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1873": {"$update": {"ARO_description": "linG is an integron-associated gene cassette, with aadA2, encoding a lincosomide nucleotidyltransferase found in Salmonella enterica. linG confers resistance to lincosomide antibiotics.", "model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_sequences": {"$update": {"sequence": {"4180": {"dna_sequence": {"fmax": "2131", "fmin": "1309", "accession": "DQ836009.1", "strand": "+", "sequence": "ATGCTTCAGCTGAAAATGATCGAACTCTTCAAGGAAGGTTGTCATGAGGATGCACGAATAATCGCGGCATTGATGTTCGGCTCATTTGCTATCGGAGAGGGTGACGAGTTCTCTGATATCGAATTCGCAGTGTTCATCCAGGATGACCATTTTGAAAATTTCGATCAGCGCTCGTGGCTTAATGCCGTAAGTCCGGTTGCTGCTTACTTTCCGGACGACTTCGGCCACCACACCGCACTTTTTGAAAACGGCATTCGCGGTGAATTCCATTTCATGCGAAAATCGGACATACCGGTCATTTCCACTTGGCAAGGCTATGGGTGGTTTCCCTCGCTTGAGGCGGCTGTTTTGTTGGACCGATCAGGAGAGTTGTCAAGGTACGCAAGCGCTCTCGTGGGCGGTCCCCCGATACGTGAAGGCGCGCCGCTGGTGGAAGGGCTTGTGTTGAACCTCATCAGCCTGATGCTCTTTGGGGCCAATCTTTTAAATCGGGGAGAGTACGCTCGCGCCTGGGCTTTGCTCAGCAAAGCACATGAAAACCTACTCAAGCTGGTTCGACTCCACGAAGGGGCAACAGACCACTGGCCGACACCTTCACGCGCGCTCGAAAAGGATATCTCGGAGGACTCGTATAATCGCTATCTGGCATGCACAAGCAGTGCAGAACCAAGAGCACTATGTGCAGCCTATCATCAAACGTGGACGTGGAGTCTCGAATTGTTCAAGAGCGTGACAGAACCTCTGAATATCGAGCTTCCGAGAACTGTAATTGCGCAGGCAAAAAGGTTGCTCAATGAGTCTGCGACGCCGCACAACAAGTAA"}, "NCBI_taxonomy": {"NCBI_taxonomy_name": "Salmonella enterica subsp. enterica serovar Stanley", "NCBI_taxonomy_id": "192953", "NCBI_taxonomy_cvterm_id": "35767"}, "protein_sequence": {"accession": "ABG65740.1", "sequence": "MLQLKMIELFKEGCHEDARIIAALMFGSFAIGEGDEFSDIEFAVFIQDDHFENFDQRSWLNAVSPVAAYFPDDFGHHTALFENGIRGEFHFMRKSDIPVISTWQGYGWFPSLEAAVLLDRSGELSRYASALVGGPPIREGAPLVEGLVLNLISLMLFGANLLNRGEYARAWALLSKAHENLLKLVRLHEGATDHWPTPSRALEKDISEDSYNRYLACTSSAEPRALCAAYHQTWTWSLELFKSVTEPLNIELPRTVIAQAKRLLNESATPHNK"}}}}}, "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1870": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1871": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1083": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1878": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1879": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "977": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "976": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "975": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "974": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "973": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "972": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "971": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "970": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1080": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "979": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "978": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "182": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "183": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "180": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "181": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "186": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "187": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "184": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "185": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2110": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2111": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "188": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "189": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2114": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2115": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2116": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "1559": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1919": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1918": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1911": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1910": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1913": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1912": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1915": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1914": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1917": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1916": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "869": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "868": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2113": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "861": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "860": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "863": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "862": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "865": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "864": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "867": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "866": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2024": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2025": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2026": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2027": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2020": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2021": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2022": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2023": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2117": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2028": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2029": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2758": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2759": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "883": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "882": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "881": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "880": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "887": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "886": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "885": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "884": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "889": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "888": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1503": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "775": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "774": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "777": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "700", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "776": {"$update": {"ARO_description": "TetX is a flavin-dependent monooxygenase conferring resistance to tetracycline antibiotics. TetX hydroxylates position 11a of the tetraketide group thus inactivating the antibiotic.", "model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "771": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "770": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "773": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "772": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "779": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "778": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "77": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "76": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "75": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "74": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "73": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "72": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "71": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "70": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "79": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "78": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1043": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1042": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1041": {"$update": {"model_description": "This model detects protein overexpression based on the presence of mutations.The detection of the protein without an associated mutation indicates that the protein is likely to be expressed at low or basal levels. The detection of the protein with the mutation indicates that the protein is likely overexpressed. This model reflects how certain proteins are functional with and without mutations. For example, efflux pump subunits and regulators are functional with mutations and without mutations. Without mutations, efflux pump subunits and regulators are usually expressed at a low level. When an efflux pump regulator has a mutation, it can cause the overexpression of the efflux pump it is responsible for regulating, leading to resistance to specific drugs. Protein overexpression models have two parameters: a curated BLASTP cutoff, and a curated set of mutations (single resistance variants, frameshift mutations, indels, etc.) shown clinically to confer resistance. This model type is a combination of the protein homolog and protein variant model. A detected hit can be categorized as Perfect, Strict, or Loose with no mutation(s) or as Strict or Loose with mutation(s).", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1040": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1047": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1046": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1045": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1044": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1049": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1048": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1681": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1680": {"$update": {"ARO_description": "MacA is a membrane fusion protein that forms an antibiotic efflux complex with MacB and TolC. macA corresponds to 1 locus in Pseudomonas aeruginosa PAO1 and 1 locus in Pseudomonas aeruginosa LESB58.", "model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1683": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1682": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1685": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1684": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1687": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1686": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1689": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1688": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1269": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1268": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "669": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "668": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "667": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1262": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "665": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "664": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "663": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "662": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1265": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1264": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "640": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1469": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1468": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "520": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "380", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1467": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1466": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1461": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1460": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1463": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1019": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1317": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1316": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1315": {"$update": {"ARO_description": "MdtC is a transporter that forms a heteromultimer complex with MdtB to form a multidrug transporter. MdtBC is part of the MdtABC-TolC efflux complex. In the absence of MdtB, MdtC can form a homomultimer complex that results in a functioning efflux complex with a narrower drug specificity. mdtC corresponds to 3 loci in Pseudomonas aeruginosa PAO1 (gene name: muxC/muxB) and 3 loci in Pseudomonas aeruginosa LESB58.", "model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "1800", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1314": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1313": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1312": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1311": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1310": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1319": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1318": {"$update": {"ARO_description": "EvgS is a sensor protein that phosphorylates the regulatory protein EvgA. evgS corresponds to 1 locus in Pseudomonas aeruginosa PAO1 and 1 locus in Pseudomonas aeruginosa LESB58.", "model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "2300", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1010": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "464": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1011": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "319": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "318": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "313": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "312": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "311": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "310": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "317": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "316": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "315": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "314": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2756": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2754": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1335": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2755": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1334": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2752": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1337": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2753": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1336": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2750": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1331": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2751": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1330": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "280", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1333": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1332": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2324": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "630": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2298": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2292": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2291": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2290": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2294": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "403": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1521": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "659": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1612": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1861": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1860": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1863": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1862": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1865": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1864": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1867": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1866": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1869": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1868": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "964": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "965": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "966": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "967": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "960": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "961": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "962": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "963": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "401": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "968": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "969": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2109": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2108": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2103": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2102": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_sequences": {"$update": {"sequence": {"4148": {"dna_sequence": {"fmax": "1454", "fmin": "0", "accession": "X52922.1", "strand": "+", "sequence": "GGCGGCGTGCTTAACACATGCAAGTCGAACGGAAAGGCCCTTTCGGGGGTACTCGAGTGGCGAACGGGTGAGTAACACGTGGGTGATCTGCCCTGCACTTTGGGATAAGCCTGGGAAACTGGGTCTAATACCGAATACACCCTGCTGGTCGCATGGCCTGGTAGGGGAAAGCTTTTGCGGTGTGGGATGGGCCCGCGGCCTATCAGCTTGTTGGTGGGGTGATGGCCTACCAAGGCGACGACGGGTAGCCGGCCTGAGAGGGTGACCGGCCACACTGGGACTGAGATACGGCCCAGACTNCTACGGGAGGCAGCAGTGGGGAATATTGCACAATGGGCGCAAGCCTGATGCAGCGACGCCGCGTGAGGGATGACGGCCTTCGGGTTGTAAACCTCTTTCAGCACAGACGAAGCGCAAGTGACGGTATGTGCAGAAGAAGGACCGGCCAACTACGTGCCAGCAGCCGCGGTAATACGTAGGGTCCGAGCGTTGTCCGGAATTACTGGGCGTAAAGAGCTCGTAGGTGGTTTGTCGCGTTGTTCGTGAAAACTCACAGCTTAACTGTGGGCGTGCGGGCGATACGGGCAGACTAGAGTACTGCAGGGGAGACTGGAATTCCTGGTGTAGCGGTGGAATGCGCAGATATCAGGAGGAACACCGGTGGCGAAGGCGGGTCTCTGGGCAGTAACTGACGCTGAGGAGCGAAAGCGTGGGGAGCGAACAGGATTAGATACCCTGGTAGTCCACGNCGTAAACGGTGGGTACTAGGTGTGGGTTTCCTTCCTTGGGATCCGTGCCGTAGCTAACGCATTAAGTANCCCGCCTGGGGAGTACGGNCGCAAGGCTAAAACTCAAAGGAATTGACGGGGGNCCGCACAAGCGGCGGAGCATGTGGATTAATTCGATGCAACGCGAAGAACCTTACCTGGGTTTGACATGCACAGGACGCCGGCAGAGATGTCGGTTCCCTTGTGGCCTGTGTGCAGGTGGTGCATGGCTGTCGTCAGCTCGTGTCGTGAGATGTTGGGTTAAGTCCCGCAACGAGCGCAACCCTTGTCTCATGTTGCCAGCACGTTATGGTGGGGACTCGTGAGAGACTGCCGGGGTCAACTCGGAGGAAGGTGGGGATGACGTCAAGTCATCATGCCCCTTATGTCCAGGGCTTCACACATGCTACAATGGCCGGTACAAAGGGCTGCGATGCCGTGAGGTGGAGCGAATCCTTTCAAAGCCGGTCTCAGTTCGGATCGGGGTCTGCAACTCGACCCCGTGAAGTCGGAGTCGCTAGTAATCGCAGATCAGCAACGCTGCGGTGAATACGTTCCCGGGCCTTGTACACACCGCCCGTCACGTCATGAAAGTCGGTAACACCCGAAGCCGGTGGCCTAACCCTTGTGGAGGGAGCCGTCGAAGGTGGGATCGGCGATTGGGACGAAGTCGTAACAAGGTAGCCG"}, "NCBI_taxonomy": {"NCBI_taxonomy_name": "Mycobacterium smegmatis", "NCBI_taxonomy_id": "1772", "NCBI_taxonomy_cvterm_id": "36871"}, "protein_sequence": {"accession": "", "sequence": ""}}}}}, "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}, "snp": {"$update": {"param_value": {"7844": "U1441C", "7843": "C1439U", "7842": "U1348C"}, "clinical": {"7844": "U1441C", "7843": "C1439U", "7842": "U1348C"}}}}}}}, "2100": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "2106": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "2105": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "2104": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1560": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "641": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "878": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "879": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "876": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "877": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "874": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "875": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "872": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "643": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "870": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "871": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2037": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2036": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2035": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1242": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2745": {"$update": {"model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2744": {"$update": {"model_param": {"$update": {"41141": {"$update": {"param_description": "This detection model parameter describes efflux pump components that are to be detected together (e.g., efflux pump subunits and regulators) using sequential model IDs, separated by commas. For example: 2685,440,1925,1305."}}}}}}, "2031": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2030": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "9": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2748": {"$update": {"model_param": {"$update": {"41141": {"$update": {"param_description": "This detection model parameter describes efflux pump components that are to be detected together (e.g., efflux pump subunits and regulators) using sequential model IDs, separated by commas. For example: 2685,440,1925,1305."}}}}}}, "2039": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "644": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "890": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "891": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "892": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "894": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "895": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "896": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "897": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "898": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "899": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "646": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1249": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "648": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1964": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1965": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1966": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1788": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1789": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "768": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "769": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1780": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1781": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "760": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "761": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "766": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "767": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "764": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "765": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1962": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1963": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1078": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1079": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1076": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1077": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1074": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1075": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1072": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1073": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1070": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1071": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1678": {"$update": {"ARO_description": "AAC(6')-Ib-cr is an aminoglycoside acetyltransferase encoded by plasmids, transposons, integrons in Enterobacteriaceae. The aac(6')-Ib-cr variant gene can induce resistance against aminoglycoside and fluoroquinolone simultaneously", "model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "ARO_category": {"$delete": ["36241"]}, "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1679": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1674": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1675": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1676": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1677": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1670": {"$update": {"model_description": "This model detects protein overexpression based on the presence of mutations.The detection of the protein without an associated mutation indicates that the protein is likely to be expressed at low or basal levels. The detection of the protein with the mutation indicates that the protein is likely overexpressed. This model reflects how certain proteins are functional with and without mutations. For example, efflux pump subunits and regulators are functional with mutations and without mutations. Without mutations, efflux pump subunits and regulators are usually expressed at a low level. When an efflux pump regulator has a mutation, it can cause the overexpression of the efflux pump it is responsible for regulating, leading to resistance to specific drugs. Protein overexpression models have two parameters: a curated BLASTP cutoff, and a curated set of mutations (single resistance variants, frameshift mutations, indels, etc.) shown clinically to confer resistance. This model type is a combination of the protein homolog and protein variant model. A detected hit can be categorized as Perfect, Strict, or Loose with no mutation(s) or as Strict or Loose with mutation(s).", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1671": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1672": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1673": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1094": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1095": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1096": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1097": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1090": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1091": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1092": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1093": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "674": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "675": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "676": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "677": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1098": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1099": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "672": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "673": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1997": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1533": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1418": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1419": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1410": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1411": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1412": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1413": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1414": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1415": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1416": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1417": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1322": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1323": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1320": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1321": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1326": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1327": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1324": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1325": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1328": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1329": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "5": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1531": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1256": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1257": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1254": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1255": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1520": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1253": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1250": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1251": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1528": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1529": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1258": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1259": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "308": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "309": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "300": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "301": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "302": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "303": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "304": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "305": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "306": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "307": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1792": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "473": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "_timestamp": "2017-08-30T14:13:02+00:00", "2478": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "470": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "471": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1898": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1899": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1894": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1895": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1896": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1897": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1890": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1891": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1892": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1893": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "959": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "958": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2134": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2135": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2132": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "2133": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "2130": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2131": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "951": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "950": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "952": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "955": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "954": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "957": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "956": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "477": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2643": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2644": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2645": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2646": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2002": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2003": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2000": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2001": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2006": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2007": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2004": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2005": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2008": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2009": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2034": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1263": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "666": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2176": {"$update": {"model_param": {"$update": {"40297": {"$update": {"param_description": "The gene order model parameter describes the relative order (5'->3') of a set of genes or other genetic elements on a chromosome, a plasmid or within an operon. Antibiotic resistance is only conferred when the detected set of genes appears in the indicated order; otherwise, no resistance phenotype is produced. This parameter is part of the gene cluster meta-model, and may be attached to detection models with the following notation: [[cvterm_id 1],[cvterm_id 2],...,[cvterm_id n]], where the cvterm_id denotes a gene-associated AMR term and an attached model id. This parameter currently (August 2017) lacks an algorithm for detection."}}}}}}, "1261": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2177": {"$update": {"model_param": {"$update": {"40297": {"$update": {"param_description": "The gene order model parameter describes the relative order (5'->3') of a set of genes or other genetic elements on a chromosome, a plasmid or within an operon. Antibiotic resistance is only conferred when the detected set of genes appears in the indicated order; otherwise, no resistance phenotype is produced. This parameter is part of the gene cluster meta-model, and may be attached to detection models with the following notation: [[cvterm_id 1],[cvterm_id 2],...,[cvterm_id n]], where the cvterm_id denotes a gene-associated AMR term and an attached model id. This parameter currently (August 2017) lacks an algorithm for detection."}}}}}}, "1799": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1798": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "719": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "718": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "717": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1267": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "715": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "714": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "713": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "712": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "711": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "710": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "661": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "716": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "505": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"40394": {"$update": {"param_type": "nonsense mutation", "param_description": "A nucleotide substitution resulting in a change from an amino acid codon to a STOP codon. Nonsense mutations truncate protein translation prematurely, resulting in a defective or completely inactive protein. In CARD, nonsense mutations may be attached to models using the notation: [wild type amino acid][position][STOP] (e.g. Q42STOP). This parameter is not currently used in detection algorithms."}}, "blastp_bit_score": {"$update": {"param_value": "900", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}, "$delete": ["40334"], "$insert": {"41345": {"param_value": {"8017": "+nt98:A", "8018": "+nt79:T"}, "param_type_id": "41345", "param_type": "insertion mutation from nucleotide sequence", "param_description": "A subtype of the insertion mutation detection model parameter. This parameter is used when a set of insertion mutations is reported in a nucleotide sequence format. Such mutations may be of variable length - possibly causing a frameshift, but not causing premature termination or a functional knockout. Mutation parameters of this type are reported in CARD with the notation: [+]nt[position]:[nucleotides]."}}}}}, "660": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2178": {"$update": {"model_param": {"$update": {"40297": {"$update": {"param_description": "The gene order model parameter describes the relative order (5'->3') of a set of genes or other genetic elements on a chromosome, a plasmid or within an operon. Antibiotic resistance is only conferred when the detected set of genes appears in the indicated order; otherwise, no resistance phenotype is produced. This parameter is part of the gene cluster meta-model, and may be attached to detection models with the following notation: [[cvterm_id 1],[cvterm_id 2],...,[cvterm_id n]], where the cvterm_id denotes a gene-associated AMR term and an attached model id. This parameter currently (August 2017) lacks an algorithm for detection."}}}}}}, "1069": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1068": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2179": {"$update": {"model_param": {"$update": {"40297": {"$update": {"param_description": "The gene order model parameter describes the relative order (5'->3') of a set of genes or other genetic elements on a chromosome, a plasmid or within an operon. Antibiotic resistance is only conferred when the detected set of genes appears in the indicated order; otherwise, no resistance phenotype is produced. This parameter is part of the gene cluster meta-model, and may be attached to detection models with the following notation: [[cvterm_id 1],[cvterm_id 2],...,[cvterm_id n]], where the cvterm_id denotes a gene-associated AMR term and an attached model id. This parameter currently (August 2017) lacks an algorithm for detection."}}}}}}, "1061": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1060": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1063": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1062": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1065": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1064": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1067": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "750", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1066": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1669": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1668": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1667": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1666": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1665": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1664": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1663": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1662": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1661": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1660": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "591": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "590": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "593": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "592": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "595": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "594": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "597": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "596": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "599": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "598": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1089": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1088": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1526": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1409": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1408": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1403": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1402": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1401": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1400": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1407": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1406": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1405": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1404": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1546": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "449": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "448": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1339": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1338": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1547": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "443": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "442": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "441": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "440": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "650", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "447": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "446": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "445": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "444": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1545": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1542": {"$update": {"ARO_description": "amrA is the efflux pump subunit of the AmrAB-OprM multidrug efflux complex. amrA corresponds to 1 locus in Pseudomonas aeruginosa PAO1 and 1 locus in Pseudomonas aeruginosa LESB58.", "model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1543": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "39": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "38": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1540": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "33": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "32": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "31": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "30": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "37": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "ARO_name": "Escherichia coli mdfA", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "700", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}, "model_name": "Escherichia coli mdfA"}}, "36": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "35": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "34": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1537": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1536": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1243": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "642": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "645": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1244": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "647": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1246": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "649": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1248": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1539": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1538": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "339": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "338": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "335": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "334": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "ARO_name": "mexX", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "337": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "336": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "331": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "330": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "333": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "332": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2032": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "8": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1464": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2119": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1462": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1889": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1888": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1887": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1886": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1885": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1884": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1883": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1882": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1881": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1880": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2121": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2123": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2122": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "2124": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "948": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "949": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "946": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "947": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "944": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1084": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "942": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "943": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "940": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "941": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2410": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2659": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2653": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2656": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2655": {"$update": {"ARO_description": "EmrE is a small multidrug transporter that functions as a homodimer and that couples the efflux of small polyaromatic cations from the cell with the import of protons down an electrochemical gradient. Confers resistance to tetraphenylphosphonium, methyl viologen, gentamicin, kanamycin, and neomycin.", "model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2654": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "133": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "132": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "131": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "130": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "137": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "136": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "135": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "134": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "139": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "138": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1354": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2019": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2018": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2015": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2014": {"$update": {"ARO_description": "PmrF is required for the synthesis and transfer of 4-amino-4-deoxy-L-arabinose (Ara4N) to Lipid A, which allows gram-negative bacteria to resist the antimicrobial activity of cationic antimicrobial peptides and antibiotics such as polymyxin. pmrF corresponds to 1 locus in Pseudomonas aeruginosa PAO1 and 1 locus in Pseudomonas aeruginosa LESB58.", "model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_name": "pmrF", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "550", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2017": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2016": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2011": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2010": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2013": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2012": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1793": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2112": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "934": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "708": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "709": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "704": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "850", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "705": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "706": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "707": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "700": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "701": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "702": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "703": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "88": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "89": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "82": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "83": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "80": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "81": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "86": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "87": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "84": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "85": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "762": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1658": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1659": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1652": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1388": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1650": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1651": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1656": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1657": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1654": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1655": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "586": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "587": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "584": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "585": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "763": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "583": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "580": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "581": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1984": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "588": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "589": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1985": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1982": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1983": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1436": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1437": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "1900", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1434": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1435": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1432": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1433": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1430": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1431": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "418": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1981": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1438": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1381": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1260": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1349": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1541": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "458": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "459": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "450": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "451": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "452": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1343": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1344": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "650", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "455": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "456": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "457": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1082": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "517": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1266": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "656": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "657": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "654": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "655": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "652": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "653": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "650": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1505": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1508": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1509": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "658": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "516": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1992": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2127": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "2126": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "2129": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "2128": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "945": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1376": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1081": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "322": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "323": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "320": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "321": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "326": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "327": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "324": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "325": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "328": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "ARO_name": "Salmonella enterica cmlA", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}, "model_name": "Salmonella enterica cmlA"}}, "329": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1340": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2331": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2332": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2333": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2335": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1594": {"$update": {"ARO_description": "vgbA inactivates streptogramin B-type antibiotics by linearizing the lactone ring on the ester bond using an elimination mechanism, thus conferring resistance to these compounds.", "model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "ARO_name": "vgbA", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1341": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1995": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2482": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "1598": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2248": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2249": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}, "$delete": ["40334"], "$insert": {"41343": {"param_value": {"8022": "-nt391:30", "8023": "-nt76:2", "8021": "-nt364:445"}, "param_type_id": "41343", "param_type": "deletion mutation from nucleotide sequence", "param_description": "A subtype of the deletion mutation detection model parameter. This parameter is used when a set of deletion mutations is reported in a nucleotide sequence format. Such mutations may be of variable length - possibly causing a frameshift, but not premature termination of functional knockout. Mutation parameters of this type are reported in the CARD with the notation: [-]nt[position]:[nucleotides]."}}}}}, "2244": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2245": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2246": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2240": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2241": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2242": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2243": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "995": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "994": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "997": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "996": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "991": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "990": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "993": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "992": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "999": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "998": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "120": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "121": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "1850", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "122": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "123": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "124": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "125": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "126": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "127": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "128": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "129": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2068": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "2069": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "2798": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2799": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2060": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2061": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2062": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2063": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2064": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2065": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2066": {"$update": {"model_description": "This model detects protein overexpression based on the presence of mutations.The detection of the protein without an associated mutation indicates that the protein is likely to be expressed at low or basal levels. The detection of the protein with the mutation indicates that the protein is likely overexpressed. This model reflects how certain proteins are functional with and without mutations. For example, efflux pump subunits and regulators are functional with mutations and without mutations. Without mutations, efflux pump subunits and regulators are usually expressed at a low level. When an efflux pump regulator has a mutation, it can cause the overexpression of the efflux pump it is responsible for regulating, leading to resistance to specific drugs. Protein overexpression models have two parameters: a curated BLASTP cutoff, and a curated set of mutations (single resistance variants, frameshift mutations, indels, etc.) shown clinically to confer resistance. This model type is a combination of the protein homolog and protein variant model. A detected hit can be categorized as Perfect, Strict, or Loose with no mutation(s) or as Strict or Loose with mutation(s).", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2795": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "1000", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2666": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2660": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2661": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1748": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1749": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1645": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1644": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1647": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "1900", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1646": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1641": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1640": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1643": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1642": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1396": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1649": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1648": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1742": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1743": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1252": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "579": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "578": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "573": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "572": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "571": {"$update": {"ARO_description": "clbA is a plasmid-encoded cfr gene found in Bacillus velezensis (Bacillus amyloliquefaciens subsp. plantarum).", "model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "570": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "577": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "576": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "575": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "574": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1209": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2800": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1208": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1421": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1420": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1423": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1422": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1425": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1424": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1427": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "1800", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1426": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1429": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1428": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2404": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "731": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "730": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2405": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "735": {"$update": {"ARO_description": "TEM-30 is an inhibitor-resistant beta-lactamase found in E. coli. Confers resistance to amoxycilllin-clavulanic acid, ticarcillin-clavulanic acid, kanamycin, neomycin, and intermediate resistance to mezlocillin and piperacillin.", "model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "734": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "737": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "736": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "739": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "738": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1359": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1358": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "469": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "468": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1353": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1352": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1351": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1350": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "461": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1356": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "463": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "608": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1273": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2158": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1519": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1518": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1515": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1514": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1517": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1516": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1511": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1510": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1513": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1512": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "280": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "582": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "357": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "356": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "355": {"$update": {"ARO_description": "TEM-1 is a broad-spectrum beta-lactamase found in many Gram-negative bacteria. Confers resistance to penicillins and first generation cephalosphorins.", "model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "354": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "353": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "352": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "351": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "350": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "359": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "358": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1033": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2323": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1447": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2321": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2326": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2325": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1446": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2329": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2328": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1445": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "675", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "289": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "288": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1444": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "281": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1443": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "283": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "282": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "285": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "284": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "287": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1442": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "600", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1441": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1116": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "263": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "262": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "261": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "260": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "267": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "266": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "265": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "264": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "286": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "269": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "268": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1562": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1563": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1564": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2192": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "650", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1565": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1566": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2259": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2257": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1567": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2251": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}, "$delete": ["40334"], "$insert": {"41343": {"param_value": {"8020": "-nt135:1", "8019": "-nt858:84"}, "param_type_id": "41343", "param_type": "deletion mutation from nucleotide sequence", "param_description": "A subtype of the deletion mutation detection model parameter. This parameter is used when a set of deletion mutations is reported in a nucleotide sequence format. Such mutations may be of variable length - possibly causing a frameshift, but not premature termination of functional knockout. Mutation parameters of this type are reported in the CARD with the notation: [-]nt[position]:[nucleotides]."}}}}}, "2476": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "988": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "989": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "982": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "983": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "980": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "981": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "986": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "850", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "987": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "984": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "985": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "115": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "114": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1790": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "116": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "111": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "110": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "113": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "112": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "119": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "118": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2785": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2787": {"$insert": {"model_param": {"blastp_bit_score": {"param_value": "650", "param_type_id": "40725", "param_type": "BLASTP bit-score", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}, "2786": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2780": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2789": {"$update": {"ARO_description": "MltA-interacting protein (mipA), is an antibiotic resistance-related outer membrane protein. Deletion of mipA increases kanamycin, nalidixic acid and streptomycin resistance.", "ARO_name": "Escherichia coli mipA"}}, "2788": {"$update": {"model_type": "protein knockout model", "model_description": "An AMR detection model for instances where the absence of a protein - due to large-scale insertion elements, large deletions, or other methods of protein knockout - confers clinical resistance to a known antibiotic. These models include reference sequences. Protein knockout models are currently in development.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}, "model_type_id": "40354"}}, "1797": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2079": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2078": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "2073": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "2072": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2071": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2070": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "2077": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "2076": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "2075": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2074": {"$update": {"model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "model_param": {"$update": {"blastn_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}}}}}}, "1796": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1035": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1389": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2679": {"$update": {"model_param": {"$update": {"41141": {"$update": {"param_description": "This detection model parameter describes efflux pump components that are to be detected together (e.g., efflux pump subunits and regulators) using sequential model IDs, separated by commas. For example: 2685,440,1925,1305."}}}}}}, "1630": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1631": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1632": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1633": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1634": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1635": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1636": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1637": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1638": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1639": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1988": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1989": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "568": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "569": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "560": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "561": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "562": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "563": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "564": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "565": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "566": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "567": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1188": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1189": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1186": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1187": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1184": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1185": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1182": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1183": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1180": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1181": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "726": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "727": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "724": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "725": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "722": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "723": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "720": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "721": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1744": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1745": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1746": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1747": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1740": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1741": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "728": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "729": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1164": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1165": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1166": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1167": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1160": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1161": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1162": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1163": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1168": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1169": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "472": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "48": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "49": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "46": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "47": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "44": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "45": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_value": "875", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "42": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "43": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "40": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "41": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1568": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1569": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1298": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1299": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1292": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1293": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1290": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1291": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1296": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1297": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1294": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1295": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1713": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1360": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1712": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2033": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1711": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "2425": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1710": {"$update": {"model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1717": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1716": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1715": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "732": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1201": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1714": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1366": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1367": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1364": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1365": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1362": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1363": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "474": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1361": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "478": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "479": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1368": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}, "1369": {"$update": {"model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "model_param": {"$update": {"blastp_bit_score": {"$update": {"param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}}}}}}, "$delete": ["2330", "857", "1134", "893", "2698"], "$insert": {"2824": {"model_id": "2824", "ARO_accession": "3004179", "model_param": {"blastn_bit_score": {"param_value": "5000", "param_type_id": "41093", "param_type": "BLASTN bit-score", "param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}, "snp": {"param_type": "single resistance variant", "param_value": {"7945": "A2063G", "7946": "A2064G"}, "clinical": {"7945": "A2063G", "7946": "A2064G"}, "param_type_id": "36301", "param_description": "A nucleotide or amino acid substitution that confers elevated resistance to antibiotic(s) relative to wild type. The most common type encoded in the CARD is an amino acid substitution gleaned from the literature with format [wild-type][position][mutation], e.g. R184Q. When present in the associated gene or protein, a single resistance variant confers resistance to an antibiotic drug or drug class. Single resistance variants are used by the protein variant and rRNA mutation models to detect antibiotic resistance from submitted sequences."}}, "ARO_description": "Point mutation in the 23S rRNA of Mycoplasma pneumoniae shown to confer resistance to erythromycin", "model_sequences": {"sequence": {"4169": {"dna_sequence": {"fmax": "2905", "fmin": "0", "accession": "NR_077056.1", "strand": "+", "sequence": "CAATAAGTTACTAAGGGCTTATGGTGGATGCCTTGGCACTAATAGGCGATGAAGGACGTGTTAACCTGCGATAAGCTTCGGGTAGGTGGTAAGAACCTCAGATCCGGAGATTTCCGAATGGAGCAATCCGGTAGTTGGAAACAGCTATCATTAATTGATGAATAAATAGTCAATTAAAGCAATACGTGGTGAAGTGAAACATCTCAGTAGCCACAGGAAAAGAAAACGAATGTGATTCCGTGTGTAGTGGCGAGCGAAAGCGGAACAGGCCAAACTTATCATTAGATAGGGGTTGTAGGGCTTGCAATGTGGACTTGAAAACGATAGAAGAAGCTGTTGGAAAGCAGCGCGCAAAAGGGTGATAGCCCCGTATTTGAAATTGTTTTCATACCTAGCGAGATCCCTGAGTAGCTCGGAAAACGTTATTTTGAGTGAATCTGCCCAGACCATTGGGTAAGCCTAAATACTAATTAGTGACCGATAGCGAAACAGTACCGTGAGGGAAAGGTGAAAAGAACCCAGAGATGGGAGTGAAATAGATTCTGAAACCATATGCCTACAACGTGTCAGAGCACATTAATGTGTGATGGCGTGCGTTTTGAAGTATGAGCCGGCGAGTTATGATAGCAAGCGTTAGTTAACCAGGAGATGGGGAGCTGTAGCGAAAGCGAGTTTTAAAAGAGCGTTTGTTTGTTATTATAGACCCGAAACGGGTTGAGCTAGTCATGAGCAGGTTGAAGGTTGAGTAACATCAACTGGAGGACCGAACCGACTCTCGTTGAAACGATAGCGGATGACTTGTGATTAGGGGTGAAATTCCAATCGAAATCCGTGATAGCTGGTTCTCGTCGAAATAGCTTTAAGGCTAGCGTGAGATCACAAATAAGTGGAGGTAAAGCTACTGAATGTATGATGGCGCCACCTAGGCGTACTGAATACAATTAAACTCTGAATGCCATTTATTTTATTCTCGCAGTCAGACAGTGGGGGATAAGCTTCATTGTCAAGAGGGGAAGAGCCCAGATCATTAAATAAGGTCCCCAAAATATACTAAGTGGAAAAGGATGTGAAAGTGCTAAAACAGCAAGGATGTTGGCTTAGAAGCAGCCATCGTTTAAAGAGTGCGTAACAGCTCACTTGTCGAGTGTTTTTGCGCCGAAGATGTAACGGGGCTAAGTATATTACCGAATTTATGGATAAGATTTATATCTTGTGGTAGACGAGCGTTGTATTGGAGTTGAAGTCAAAGCGTGAGCATTGGTGGATCCAATACAAGTGAGAATGCCGGCATGAGTAACGCTTGGGAGTGAGAATCTCCCAAACCGATTGACTAAGGTTTCCTGGACCAGGGTCGTCCTTCCAGGGTTAGTCTGGACCTAAGCTGAGGCTGAAAAGCGTAGGCGATGGACAACAGGTTAATATTCCTGTACTTACAGTTAGACTGATGGAGTGACAAAGAAGGTTTTCCACCCCCATAATTGGATTTGGGGATAAATCATAAGGTGGTACAATAGGCAAATCCGTTGTGCATAACATTGAGTGATGATGTCGAGTGAATGAGTGATCAAGTAGCGAAGGTGGTATTAATCATGCTTTCAAGAAAAGCTTCTAGGGTTAATCTAGCTGTAACCAGTACCGAGAACGAACACACGTAGTCAAGGAGAGGATCCTAAGGTTAGCGAGTGAACTATAGCCAAGGAACTCTGCAAATTAACCCCGTAAGTTAGCGAGAAGGGGTGCTTATGTAAAAGTAAGCCGCAGTGAAGAACGAGGGGGGACTGTTTAACTAAAACACAACTCTATGCCAAACCGTAAGGTGATGTATATGGGGTGACACCTGCCCAGTGCTGGAAGGTTAAAGAAGGAGGTTAGCGCAAGCGAAGCTTTTAACTGAAGCCCCAGTGAACGGCGGCCGTAACTATAACGGTCCTAAGGTAGCGAAATTCCTAGTCGGGTAAATTCCGTCCCGCTTGAATGGTGTAACCATCTCTTGACTGTCTCGGCTATAGACTCGGTGAAATCCAGGTACGGGTGAAGACACCCGTTAGGCGCAACGGGACGGAAAGACCCCGTGAAGCTTTACTGTAGCTTAATATTGATCAGGACATTATCATGTAGAGAATAGGTAGGAGCAATCGATGCAAGTTCGCTAGGACTTGTTGATGCGAAAGGTGGAATACTACCCTTGGTTGTGTGCTGTTCTAATTGGTAACTGTTATCCAGTTTCAAGACAGTGTTAGGTGGGCAGTTTGACTGGGGCGGTCGCCTCCTAAAAGGTAACGGAGGCGTACAAAGGTACCTTCAGTACGGTTGGAAATCGTATGTAGAGTGTAATGGTGTAAGGGTGCTTGACTGTGAGACATACAGGTCGAACAGGTGAGAAATCAGGTCATAGTGATCCGGTGGTCCAGTATGGAATGGCCATCGCTCAACGGATAAAAGCTACTCCGGGGATAACAGGCTGATACTGCCCAAGAGTTCATATCGACGGCAGTGTTTGGCACCTCGATGTCGACTCATCTCATCCTCGAGCTGAAGCAGGTTCGAAGGGTTCGGCTGTTCGCCGATTAAAGAGATACGTGAGTTGGGTTCAAACCGTCGTGAGACAGGTTGGTCCCTATCTATTGTGCCCGTAGGAAGATTGAAGAGTGTTGCTTCTAGTACGAGAGGACCGAAGCGAGGACACCTCTTATGCTCCAGTTGTAGCGCCAGCTGCACCGCTGGGTAGTAACGTGTCTATTAGATAAACGCTGAAAGCATCTAAGTGTGAAACTATCTCAAAGATTAATCTTCCCATTTCGCAAGAAAGTAAGAGCCGTCAAAGACGATGACGTTGATAGGTTACAGGTGTAAGCATAGTGATATGTTGAGCTGAGTAATACTAATTGCTCGAGGACTTATTGGA"}, "NCBI_taxonomy": {"NCBI_taxonomy_name": "Mycoplasma pneumoniae", "NCBI_taxonomy_id": "2104", "NCBI_taxonomy_cvterm_id": "41332"}, "protein_sequence": {"accession": "", "sequence": ""}}}}, "ARO_category": {"35950": {"category_aro_name": "antibiotic resistant gene variant or mutant", "category_aro_cvterm_id": "35950", "category_aro_accession": "0000031", "category_aro_description": "Resistance to antibiotics is often conferred by single nucleotide polymorphisms (SNPs) and other mutations in target genes."}, "36454": {"category_aro_name": "determinant of macrolide resistance", "category_aro_cvterm_id": "36454", "category_aro_accession": "3000315", "category_aro_description": "Enzymes, other proteins or other gene products shown clinically to confer resistance to macrolide antibiotics."}}, "ARO_name": "Mycoplasma pneumoniae 23S rRNA mutation conferring resistance to erythromycin", "model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "ARO_id": "41331", "model_name": "Mycoplasma pneumoniae 23S rRNA mutation conferring resistance to erythromycin", "model_type_id": "40295"}, "2808": {"model_id": "2808", "ARO_accession": "3004161", "model_param": {"blastn_bit_score": {"param_value": "5000", "param_type_id": "41093", "param_type": "BLASTN bit-score", "param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}, "snp": {"param_type": "single resistance variant", "param_value": {"7900": "G2281A", "7901": "G2294A", "7902": "G2295A"}, "clinical": {"7900": "G2281A", "7901": "G2294A", "7902": "G2295A"}, "param_type_id": "36301", "param_description": "A nucleotide or amino acid substitution that confers elevated resistance to antibiotic(s) relative to wild type. The most common type encoded in the CARD is an amino acid substitution gleaned from the literature with format [wild-type][position][mutation], e.g. R184Q. When present in the associated gene or protein, a single resistance variant confers resistance to an antibiotic drug or drug class. Single resistance variants are used by the protein variant and rRNA mutation models to detect antibiotic resistance from submitted sequences."}}, "ARO_description": "Point mutation in the 23S rRNA of Propionibacteria shown clinically to confer resistance to macrolides", "model_sequences": {"sequence": {"4155": {"dna_sequence": {"fmax": "4984", "fmin": "1849", "accession": "Y10819.1", "strand": "+", "sequence": "GTGACAAGCTACTAAGTGCGATCGGTGGATGCCTAGGCACCAAGAGCCGATGAAGGACGTTGTAACCTGCGATAAGCCCTGGGGAGCTGGTAAACGAGCTTTGATCCGGGGATGTCCGAATGGGGAAACCTCGAAGGTGACCAGTTTAGCTACTGGCGACCGCCGCCTGAATGTATAGGGCGGTTGGAGGGAACGTGGGGAAGTGAAACATCTTAGTACCCACAGGAAGAGRAAACAACCGTGAWTCCGTGAATATTGGCGAGCGAAAGCGGAAGAGGCCAAACCGGAGTGTGTGATAGCCGGCAGGTGTTGCATGTGCGGGGTTGTGGGAAGCGTTTTGACTGAACTGCCGTGAGGTCGGAGAGTGATAAAGGATTGATGAAGCAGAAGCGTCTGGGAAGGCGCGGCATAGATGGTGATACCCCTGTATGCGTAAGTTGATCTCTCTCTTAATGTTTTCCCAAGTAGTACGGAACCCCTGAAATYCCGTACGAATCTGGCGGACCACCCGTTAAGCCTAAATACBCCTTGGTGACCGATMGCGGACAAGTACCCGTGAGGGAAAGGGTGAAAATGTACCCCCGGGAGGGGAGTGAAATAGTMCCTGAAACCGATCGCATACAATCCGTCGGAGCCTGCCCTTGTGGTGGGTGACGGCGTGCCTTTTGRAAGAATGAGCCTGCGAGTTAGTGGTGTGTGGCGAGGTTAACCCGTGTGGGGAAGCCGTAGCGAAAGCGAGTCCGAATAGGGCGTTTGAGTCGCATGCTCTAGACCCGAAGCGGTGTGATCTATCCATGGCCAGGGTGAAGCGACGGTAAGACGTCGTGGAGGCCCGAACCCACCAGGGTTGCAAACCTGGGGGATGAGCTGTGGATAGGGGTGAAAGGCCAATCAAACACCGTGATAGCTGGTTCTCCCCGAAATGCATTTAGGTGCAGCGTCATGTGTTTCTTGTCGGAGGTAGAGCACTGGATGGTCTAGGGGGCTTACCAGCTTACCGAAATCAGCCAAACTCCGAATGCCGACAAGTGAGAGCATGGCAGTGAGACGGCGGGGGATAAGCTTCGTCGTCGAGAGGGAAACAGCCCAGATCATCAGCTAAGGCCCCTAAGTGGTGACTAAGTGGAAAAGGACGTGGAGTTGCGGAGACAACCAGGAGGTTGGCTTGGAAGCAGCCATCCTTGAAAGAGTGCGTAATAGCTCACTGGTCAAGTGATTCTGCACCGACAATTTAGCGGGGCTCAAGTCATCCGCCGAAGCTGTGGCATCTACGCGTGTATCCGGCATCCTTTGGGGTGTCCAGGTGCGTGGATGGGTAGGGGAGCGTTGTGTGTGCGTTGAAGCGGCGGGGTGACCCGGTCGTGGAGTGCACGCAAGTGAGAATGCAGGCATGAGTAGCGTATGACGGGTGAGAAACCCGTCCGCCGAATATCCAAGGGTTCCAGGGTCAAGCTAATCTGCCCTGGGTGAGTCGGGTCCTAAGGCGAGGCCGACAGGCGTAGTCGATGGACAACGGGTTGATATTCCCGTACCGGCGCGAGAACGATCCTGCCGAGGTGAGTGATGCTAAGCATGCAAGGCGGTCGTGGGGGCTTCGGTTCCCTGATCGTTGAGTCTGTAACCCGATCTTGTAGTAGGCAAGCTGCGGAGGGACGCAGGAAGGTAGTCTGGCACCGTATTGGTTTGCGGTGTTAAGCCTGTAGGGTGTCTGGCCAGGTAAATCCGGTCGGACGTGTGCCTGAGAGGTGATGAGTGGTGCCACTTTTGTGGTACGTATCCGGATGATCCTATGCTGCCTAGAAAATCTTCGTGAGCGAGTTCTCGAGCTGCCCGTACCCCAAACCGACACTGGTGGATAGGTAGAGAATACCAAGGCGATCGAGATAATCATGGTGAAGGAACTCGGCAAAATCCTCCCGTAACTTCGGAATAAGGGAGACTGGAGGCGTGACGGCAGTTTACTTGTCGGTGCGTCGATAGTCGCAGAGAATAGGCCCAAGCGACTGCTTACTAAAAGCACAGGTCCGTGCTAAGTCGAAAGACGATGTATACGGACTGACTCCTGCCCGGTGCTCKGAAGGTTAAGGGGACGTGTTAGCACTTTTGTGCGAGGCACTGAACTTAAGCCCCAGTAAACGGCGGTGGTAACTATAACCATCCTAAGGKAGCGAAATTCCTTGTCGGGKAAGTTCCGRCCTGSACGAATGGAGTAACGACTTGGGCGCTGTCTCCACCATGAACTCGGCGAAATTGCATTACGAGTAAAGATGCTCGTTACGCGCACAGGGACGGAAAGVNCCCGGGACCTTTACTATAGTTTGGTATTGGTGATCGGTACGACTTGTGTAGGATAGGTGGGAGACTTTGAAGCGGTCACGCTAGTGATTGTGGAGTCATTGTTGAAATACCACTCTGGTCGTTCTGGTTATCTAACCTAGGTCCGTGATCCGGATCAGGGACAGTGCCTGATGGGTAGTTTGACTGGGGCGGTCGCCTCCMAAAAGGTAACGGAGGCGCCCAAAGGTTCCCTCAGCCTGGTTGGTAATCAGGTGTTGAGTGTAAGTGCACAAGGGAGCTTGACTGTGAGACAGACATGTCGAGCAGGGACGAAAGTCGGGACTAGTGATCYTCTGGTGGATTGTGGAATCGCCAGAACTCAACGGATAAAAGGTACCCCGGGGATAACAGGCTGATCTTTCCCGAGCGCTCACAGCGACGGAATGGNTTGGCACCTCGATGTCGGCTCGTCGCATCCTGGGGCTGGAGTCGGTCCCAAGGGTTGGGCTGTTCGCCCATTAAAGCGGCACGCGARGCTGGGGKTAAGAACGTCGTGAGACAGTTCCGGKCCCTATACCGCTGCSCGTMGKATCTTGAGAGGGCTGTCCTTAGTACGCAAGGACCGGGACGGACCAACCTCTGGTGTGCCAGTTGTTCCACCAGGAGCATGGCTGGTTGGCTACGTTGGGGAGTGATAACCGCTGAAAGCATCTAAGTGGGAAGCACGCTTCAAGATGAGGGTTCCTGCACAGTTAATGTGGTAAGGCCCCCGGTAGACCACCGGGTGATAGGTCGGATGTGGAAGCATGGTGACATGTGGAGCTGACCGATACTAAGTGGCCGAGGGCTTGTCCCACA"}, "NCBI_taxonomy": {"NCBI_taxonomy_name": "Propionibacterium freudenreichii", "NCBI_taxonomy_id": "1744", "NCBI_taxonomy_cvterm_id": "41309"}, "protein_sequence": {"accession": "", "sequence": ""}}}}, "ARO_category": {"35950": {"category_aro_name": "antibiotic resistant gene variant or mutant", "category_aro_cvterm_id": "35950", "category_aro_accession": "0000031", "category_aro_description": "Resistance to antibiotics is often conferred by single nucleotide polymorphisms (SNPs) and other mutations in target genes."}, "36454": {"category_aro_name": "determinant of macrolide resistance", "category_aro_cvterm_id": "36454", "category_aro_accession": "3000315", "category_aro_description": "Enzymes, other proteins or other gene products shown clinically to confer resistance to macrolide antibiotics."}}, "ARO_name": "Propionibacteria 23S rRNA with mutation conferring resistance to macrolide antibiotics", "model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "ARO_id": "41308", "model_name": "Propionibacteria 23S rRNA with mutation conferring resistance to macrolide antibiotics", "model_type_id": "40295"}, "2828": {"model_id": "2828", "ARO_accession": "3004185", "model_param": {"blastp_bit_score": {"param_value": "1000", "param_type_id": "40725", "param_type": "BLASTP bit-score", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}, "ARO_description": "mecD is a PBP2a variant identified on a genomic resistance island in Macrococcus caseolyticus. MecD confers resistance to methicillin and other beta-lactam antibiotics through the production of an alternative low-affinity PBP. First described by Schwendener et al. 2017 and identified from canine and bovine sources.", "model_sequences": {"sequence": {"4173": {"dna_sequence": {"fmax": "17474", "fmin": "15437", "accession": "KY013611.1", "strand": "+", "sequence": "ATGAAGAATATAAAAGTTAAAATATTAATAGTTTGTAGCCTATGTCTTATATCATTTTTCTTATATAATTTATTGAAAGAAAATGAAATTGATAAAATATTTTCTTCAATTGAAAATCGAAATGTAGATGAAATAAATGAGAATATTACTTTTTTATCAAGAAATACTTTTAGTAAAAAACAAAGATACGATAGAATGAATCATATTGATAATTCTCTAGGCATAAAAAAAGTAAATATAACTGATATAAAGTTATTAGAAGAAATTGTAGACACTCGAAAATATAGTGCTAATATGCACTATGATTCTAAATTCGGTAAATTCACAAAAAAAGGTTATTTTGAGTTTGAAAAAAATGGTGAGAGTAAACGTTGGGAATTGAACTGGACACCAGAGGTTATAATTCCAGGACTTACAGCAACTAATGAAGTACGAGTAGAAGAATTGAAATCAAGTAGAGGTGAAATTGTAGATAGAAATGGAATTCCTTTAGCGATAGATGGTGAACATTATCAGGTAGGAATTGATCCTAAAAATTATAATAAGAAGGATAGTAAACAAATAGCTAAGTTATTAAATATAAATGAGAGTACTTTAAAGAACAAATTAAAACAATCATGGGTTAAAGATGGAGTATTTGTTCCAATCAAATCATATGTAGAATTAAGTGATGAAATTAAAAATAAAATACCTGAATACGGATTATCTGTGAATAAAATAAAAGGTAGAACTTACCCATTAAAAGAAGCAAGTGCTCATTTGTTAGGTTATATTGGTGAAATAAATGCAGATGAGCTTAATGATCCAAAATTTAAAGGATATGATTCTCATTCAATTGTTGGTAAAACTGGTATTGAATATATGTACGATAAAGAATTACAAAACAGAGATGGTTTAATAGTATATATTACTGATGATGATGGTTTAACAGATTCAAAAGAAATATTAGTTCATAAGAAACCTAAAAATGGAAAAAAGATAGTTTTATCGATAGATAGTAGAGTTCAAAATAGTATTTACAATCATTTAAAAGATGATAATGGATCTGGTACTGCTATGAATCCCAAAACAGGAGAATTATTAGCTTTAGTAAGCTATCCATCATTTAATCCTTATGATTTTATGTTTGGTATTTCAAATAAAAAATATCAGGCCTTATTAAATGATAAAAAGGCTCCGCTTTTAAATAAATTTCAAGAATTAACTTCTCCAGGTTCTACTCAAAAATTGCTAACTTCTATAATAGGTTTAAATAATGGGGTTATAAATGAAAGCAAAAGTTATGAAATAAATGGTAAAGGATGGAGAAAAGATGGGAGCTGGGGAGGATATAAAGTAACAAGATTTGAAGTTGTCAATGGACGAATTGATTTAGAGAAAGCTATAGCACATTCAGACAATATCTTTTTTGCTAGAACAACCCTAGAAATGGGTGGGAAAAAATTTGTAAGGGGAATGAAGGATTTAGGTGTAGGAGAGGAAACGCCTTCTGATTACCCTGTTCGGACTGGCCAAATAGCGAATAAAATTAATCTAGAAAGAAATTTGAATAATGATATATTGTTAGCAGACTCAGGTTATGGACAAGGTGAAATATTGGTGAATCCAATTCATATTTTATCTATATATAGTTCATTAGTAAATGAAGGGAATATGATGGCACCTAAATTAAATATGGAACATAAAAGCAAAGTATGGAAAAAACATATTACATCACAAAAAAATATTGATATATTAACAAGTAGTATGAGAAAAGTTGTTACAGGCACTCATAAATTAGATACCGAAAGAAATTATGCTAACTTTGCTGGAAAAACTGGTACAGCAGAATTAAAAATGACTCAGAATGAAGGATTAGGTACACAAATTGGATGGTTTGTGGGTTATGATCAACAAAATCCAAATATGATGTTAGCTATAAATGTTAAAAATGTTGAAGACAAAGGTATGTCTAGCTATAATGCACAAAAATTTGCGCAAGTAATGGATGATTTATATGAACATGGAGCTAGGACTTATGAACCAGACTCAGAATAA"}, "NCBI_taxonomy": {"NCBI_taxonomy_name": "Macrococcus caseolyticus", "NCBI_taxonomy_id": "69966", "NCBI_taxonomy_cvterm_id": "40025"}, "protein_sequence": {"accession": "AQX82857.1", "sequence": "MKNIKVKILIVCSLCLISFFLYNLLKENEIDKIFSSIENRNVDEINENITFLSRNTFSKKQRYDRMNHIDNSLGIKKVNITDIKLLEEIVDTRKYSANMHYDSKFGKFTKKGYFEFEKNGESKRWELNWTPEVIIPGLTATNEVRVEELKSSRGEIVDRNGIPLAIDGEHYQVGIDPKNYNKKDSKQIAKLLNINESTLKNKLKQSWVKDGVFVPIKSYVELSDEIKNKIPEYGLSVNKIKGRTYPLKEASAHLLGYIGEINADELNDPKFKGYDSHSIVGKTGIEYMYDKELQNRDGLIVYITDDDGLTDSKEILVHKKPKNGKKIVLSIDSRVQNSIYNHLKDDNGSGTAMNPKTGELLALVSYPSFNPYDFMFGISNKKYQALLNDKKAPLLNKFQELTSPGSTQKLLTSIIGLNNGVINESKSYEINGKGWRKDGSWGGYKVTRFEVVNGRIDLEKAIAHSDNIFFARTTLEMGGKKFVRGMKDLGVGEETPSDYPVRTGQIANKINLERNLNNDILLADSGYGQGEILVNPIHILSIYSSLVNEGNMMAPKLNMEHKSKVWKKHITSQKNIDILTSSMRKVVTGTHKLDTERNYANFAGKTGTAELKMTQNEGLGTQIGWFVGYDQQNPNMMLAINVKNVEDKGMSSYNAQKFAQVMDDLYEHGARTYEPDSE"}}}}, "ARO_category": {"35929": {"category_aro_name": "antibiotic resistance gene cluster, cassette, or operon", "category_aro_cvterm_id": "35929", "category_aro_accession": "0000010", "category_aro_description": "Clusters of antibiotic resistance genes. May be regulated by a shared promoter or repressor."}, "36268": {"category_aro_name": "determinant of beta-lactam resistance", "category_aro_cvterm_id": "36268", "category_aro_accession": "3000129", "category_aro_description": "Enzymes, other proteins or other gene products shown clinically to confer resistance to beta-lactam antibiotics."}, "36520": {"category_aro_name": "antibiotic target replacement protein", "category_aro_cvterm_id": "36520", "category_aro_accession": "3000381", "category_aro_description": "Alternate proteins that have the same functions as other antibiotic target proteins, but are structurally different and thus resistant to antibiotics. These can replace the activity of other antibiotic-sensitive proteins in the presence of antibiotics."}}, "ARO_name": "mecD", "model_type": "protein homolog model", "model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "ARO_id": "41338", "model_name": "mecD", "model_type_id": "40292"}, "2829": {"model_id": "2829", "ARO_accession": "3002855", "model_param": {"blastp_bit_score": {"param_value": "300", "param_type_id": "40725", "param_type": "BLASTP bit-score", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}, "ARO_description": "dfrA2d is an integron-encoded dihydrofolate reductase found in Enterobacteriaceae", "model_sequences": {"sequence": {"4174": {"dna_sequence": {"fmax": "362", "fmin": "125", "accession": "AY973253.1", "strand": "+", "sequence": "ATGAATGAAGGAAAAAATGAGGTCAGTACTTCAGCTGCTGGCCGGTTCGCATTCCCATCAAACGCCACGTTTGCCTTGGGGGATCGCGTACGCAAGAAGTCTGGCGCTGCTTGGCAGGGGCGCATTGTCGGGTGGTACTGCACAACACTTACCCCTGAAGGCTACGCCGTCGAGTCCGAATCTCACCCAGGCTCAGTCCAGATTTATCCCATGACTGCGCTTGAACGGGTGGCCTGA"}, "NCBI_taxonomy": {"NCBI_taxonomy_name": "Escherichia coli", "NCBI_taxonomy_id": "562", "NCBI_taxonomy_cvterm_id": "35914"}, "protein_sequence": {"accession": "AAX84553.1", "sequence": "MNEGKNEVSTSAAGRFAFPSNATFALGDRVRKKSGAAWQGRIVGWYCTTLTPEGYAVESESHPGSVQIYPMTALERVA"}}}}, "ARO_category": {"36520": {"category_aro_name": "antibiotic target replacement protein", "category_aro_cvterm_id": "36520", "category_aro_accession": "3000381", "category_aro_description": "Alternate proteins that have the same functions as other antibiotic target proteins, but are structurally different and thus resistant to antibiotics. These can replace the activity of other antibiotic-sensitive proteins in the presence of antibiotics."}, "40990": {"category_aro_name": "determinant of diaminopyrimidine resistance", "category_aro_cvterm_id": "40990", "category_aro_accession": "3004028", "category_aro_description": "Enzymes, other proteins or other gene products shown clinically to confer resistance to diaminopyrimidine (incl. trimethoprim) antibiotics."}}, "ARO_name": "dfrA2d", "model_type": "protein homolog model", "model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "ARO_id": "39289", "model_name": "dfrA2d", "model_type_id": "40292"}, "2803": {"model_id": "2803", "ARO_accession": "3004153", "model_param": {"41345": {"param_value": {"8079": "+nt217:CACGAGCAC", "8080": "+nt372:T", "8082": "-nt472:C"}, "param_type_id": "41345", "param_type": "insertion mutation from nucleotide sequence", "param_description": "A subtype of the insertion mutation detection model parameter. This parameter is used when a set of insertion mutations is reported in a nucleotide sequence format. Such mutations may be of variable length - possibly causing a frameshift, but not causing premature termination or a functional knockout. Mutation parameters of this type are reported in CARD with the notation: [+]nt[position]:[nucleotides]."}, "snp": {"param_type": "single resistance variant", "param_value": {"7852": "H75N", "7862": "S105P", "7853": "T22A", "7855": "Y36C", "7867": "I211V", "7866": "H207R", "7865": "C161T", "7864": "F152V", "7863": "R126Q", "7857": "V77F", "7854": "T202A", "7860": "G91R", "7851": "W101R", "7858": "W83C", "7869": "R235P", "7868": "P224L"}, "clinical": {"7852": "H75N", "7862": "S105P", "7853": "T22A", "7855": "Y36C", "7867": "I211V", "7866": "H207R", "7865": "C161T", "7864": "F152V", "7863": "R126Q", "7857": "V77F", "7854": "T202A", "7860": "G91R", "7851": "W101R", "7858": "W83C", "7869": "R235P", "7868": "P224L"}, "param_type_id": "36301", "param_description": "A nucleotide or amino acid substitution that confers elevated resistance to antibiotic(s) relative to wild type. The most common type encoded in the CARD is an amino acid substitution gleaned from the literature with format [wild-type][position][mutation], e.g. R184Q. When present in the associated gene or protein, a single resistance variant confers resistance to an antibiotic drug or drug class. Single resistance variants are used by the protein variant and rRNA mutation models to detect antibiotic resistance from submitted sequences."}, "40394": {"param_value": {"7859": "W83STOP", "7856": "G76STOP", "7861": "W98STOP"}, "param_type_id": "40394", "param_type": "nonsense mutation", "param_description": "A nucleotide substitution resulting in a change from an amino acid codon to a STOP codon. Nonsense mutations truncate protein translation prematurely, resulting in a defective or completely inactive protein. In CARD, nonsense mutations may be attached to models using the notation: [wild type amino acid][position][STOP] (e.g. Q42STOP). This parameter is not currently used in detection algorithms."}, "41343": {"param_value": {"8077": "-nt111:T", "8076": "-nt260:C"}, "param_type_id": "41343", "param_type": "deletion mutation from nucleotide sequence", "param_description": "A subtype of the deletion mutation detection model parameter. This parameter is used when a set of deletion mutations is reported in a nucleotide sequence format. Such mutations may be of variable length - possibly causing a frameshift, but not premature termination of functional knockout. Mutation parameters of this type are reported in the CARD with the notation: [-]nt[position]:[nucleotides]."}, "blastp_bit_score": {"param_value": "300", "param_type_id": "40725", "param_type": "BLASTP bit-score", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}, "ARO_description": "Point mutations in the thymidylate synthetase thyA gene shown clinically to confer resistance to para-aminosalicylic acid. Loss-of-function mutations in thyA disrupt the catalytic activity and substrate-binding affinity, thus conferring resistance.", "model_sequences": {"sequence": {"4150": {"dna_sequence": {"fmax": "3074471", "fmin": "3073679", "accession": "NC_000962.3", "strand": "-", "sequence": "TCATACCGCGACTGGAGCTTTGATCGCCGGATGCGGATCGTAGTTCTTCACAACGATGTCTTCATAGGTGTACTCGAAGATTGAATCCCGGTCGGCTAGAAGTAGTTTCGGATATGGCCGCGGCTCGCGGCTGAGCTGCAGCCGTACTTGCTCGACGTGATTGTCGTAGATGTGGCAGTCGCCACCGGTCCAGATGAACTCGCCGACCGACAAGCCGGCCTGGGCGGCCATCATGTGGGTGAGCAACGCATAGCTGGCGATGTTGAACGGCACACCCAGAAACAGGTCGGCGCTGCGTTGGTAGAGCTGACAGCTCAGCCGGCCATCGGCGACGTAGAACTGGAAGAACGCATGACAGGGCGGCAGCGCCATCCGCTCGATTTCGCCGACGTTCCAGGCCGACACGATGATGCGCCGGGAATCGGGATCGGTGCGCAGCAAATCCAGCGCCGCGCTGATCTGGTCGATGTGCTCACCGGATGGAGCCGGCCACGATCGCCATTGTACACCGTAGATCGGCCCGAGTTCGCCTGTATCACTTGCCCATTCGTCCCAGATGGTGACTCCGTGCTCGTGCAGCCAACCGATATTGGAATCGCCGCGCAAAAACCACAGCAGCTCGTAGGCTACCGATTTGAAATGGACTTTCTTGGTAGTGAGCAGCGGGAAACCGGCCGACAAATCATAGCGCATCTGCTGGCCGAACAGGCTGCGGGTTCCGGTGCCGGTGCGGTCGGATTTGGGCGTACCCGTTTCGAGCACGAAGCGCAGCAGGTCCTCGTATGGCGTCAC"}, "NCBI_taxonomy": {"NCBI_taxonomy_name": "Mycobacterium tuberculosis H37Rv", "NCBI_taxonomy_id": "83332", "NCBI_taxonomy_cvterm_id": "39507"}, "protein_sequence": {"accession": "NP_217280.1", "sequence": "MTPYEDLLRFVLETGTPKSDRTGTGTRSLFGQQMRYDLSAGFPLLTTKKVHFKSVAYELLWFLRGDSNIGWLHEHGVTIWDEWASDTGELGPIYGVQWRSWPAPSGEHIDQISAALDLLRTDPDSRRIIVSAWNVGEIERMALPPCHAFFQFYVADGRLSCQLYQRSADLFLGVPFNIASYALLTHMMAAQAGLSVGEFIWTGGDCHIYDNHVEQVRLQLSREPRPYPKLLLADRDSIFEYTYEDIVVKNYDPHPAIKAPVAV"}}}}, "ARO_category": {"35950": {"category_aro_name": "antibiotic resistant gene variant or mutant", "category_aro_cvterm_id": "35950", "category_aro_accession": "0000031", "category_aro_description": "Resistance to antibiotics is often conferred by single nucleotide polymorphisms (SNPs) and other mutations in target genes."}, "41298": {"category_aro_name": "determinant of aminosalicylate resistance", "category_aro_cvterm_id": "41298", "category_aro_accession": "3004151", "category_aro_description": "Enzymes, other proteins or other gene products shown clinically to confer resistance to aminosalicylates, esp. para-aminosalicylic acid, commonly used to treat Mycobacterium infections."}}, "ARO_name": "Mycobacterium tuberculosis thyA with mutation conferring resistance to para-aminosalicylic acid", "model_type": "protein variant model", "model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "ARO_id": "41300", "model_name": "Mycobacterium tuberculosis thyA with mutation conferring resistance to para-aminosalicylic acid", "model_type_id": "40293"}, "2802": {"model_id": "2802", "ARO_accession": "3004131", "model_param": {"blastn_bit_score": {"param_value": "5000", "param_type_id": "41093", "param_type": "BLASTN bit-score", "param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}, "snp": {"param_type": "single resistance variant", "param_value": {"7845": "U754A", "7847": "G2057A", "7849": "A2058U", "7850": "C2611U", "7848": "A2058G"}, "clinical": {"7845": "U754A", "7847": "G2057A", "7849": "A2058U", "7850": "C2611U", "7848": "A2058G"}, "param_type_id": "36301", "param_description": "A nucleotide or amino acid substitution that confers elevated resistance to antibiotic(s) relative to wild type. The most common type encoded in the CARD is an amino acid substitution gleaned from the literature with format [wild-type][position][mutation], e.g. R184Q. When present in the associated gene or protein, a single resistance variant confers resistance to an antibiotic drug or drug class. Single resistance variants are used by the protein variant and rRNA mutation models to detect antibiotic resistance from submitted sequences."}}, "ARO_description": "Point mutation in the 23S rRNA of Escherichia coli shown clinically to confer resistance to the macrolide-class antibiotics erythromycin and telithromycin.", "model_sequences": {"sequence": {"4149": {"dna_sequence": {"fmax": "240063", "fmin": "237159", "accession": "AE014075.1", "strand": "+", "sequence": "GGTTAAGCGACTAAGCGTACACGGTGGATGCCCTGGCAGTCAGAGGCGATGAAGGACGTGCTAATCTGCGATAAGCGTCGGTAAGGTGATATGAACCGTTATAACCGGCGATTTCCGAATGGGGAAACCCAGTGTGTTTCGACACACTATCATTAACTGAATCCATAGGTTAATGAGGCGAACCGGGGGAACTGAAACATCTAAGTACCCCGAGGAAAAGAAATCAACCGAGATTCCCCCAGTAGCGGCGAGCGAACGGGGAGGAGCCCAGAGCCTGAATCAGTGTGTGTGTTAGTGGAAGCGTCTGGAAAGGCGCGCGATACAGGGTGACAGCCCCGTACACAAAAATGCACATGCTGTGAGCTCGATGAGTAGGGCGGGACACGTGGTATCCTGTCTGAATATGGGGGGACCATCCTCCAAGGCTAAATACTCCTGACTGACCGATAGTGAACCAGTACCGTGAGGGAAAGGCGAAAAGAACCCCGGCGAGGGGAGTGAAAAAGAACCTGAAACCGTGTACGTACAAGCAGTGGGAGCATGCTTAGGCGTGTGACTGCGTACCTTTTGTATAATGGGTCAGCGACTTATATTCTGTAGCAAGGTTAACCGAATAGGGGAGCCGAAGGGAAACCGAGTCTTAACTGGGCGTTAAGTTGCAGGGTATAGACCCGAAACCCGGTGATCTAGCCATGGGCAGGTTGAAGGTTGGGTAACACTAACTGGAGGACCGAACCGACTAATGTTGAAAAATTAGCGGATGACTTGTGGCTGGGGGTGAAAGGCCAATCAAACCGGGAGATAGCTGGTTCTCCCCGAAAGCTATTTAGGTAGCGCCTCGTGAACTCATCTCCGGGGGTAGAGCACTGTTTCGGCAAGGGGGTCATCCCGACTTACCAACCCGATGCAAACTGCGAATACCGGAGAATGTTATCACGGGAGACACACGGCGGGTGCTAACGTCCGTCGNGAAGAGGGAAACAACCCAGACCGCCAGCTAAGGTCCCAAAGTCATGGTTAAGTGGGAAACGATGTGGGAAGGCCCAGACAGCCAGGATGTTGGCTTAGAAGCAGCCATCATTTAAAGAAAGCGTAATAGCTCACTGGTCGAGTCGGCCTGCGCGGAAGATGTAACGGGGCTAAACCATGCACCGAAGCTGCGGCAGCGACGCTTATGCGTTGTTGGGTAGGGGAGCGTTCTGTAAGCCTGTGAAGGTGTACTGTGAGGTATGCTGGAGGTATCAGAAGTGCGAATGCTGACATAAGTAACGATAAAGCGGGTGAAAAGCCCGCTCGCCGGAAGACCAAGGGTTCCTGTCCAACGTTAATCGGGGCAGGGTGAGTCGACCCCTAAGGCGAGGCCGAAAGGCGTAGTCGATGGGAAACAGGTTAATATTCCTGTACTTGGTGTTACTGCGAAGGGGGGACGGAGAAGGCTATGTTGGCCGGGCGACGGTTGTCCCGGTTTAAGCGTGTAGGCTGGTTTTCCAGGCAAATCCGGAAAATCAAGGCTGAGGCGTGATGACGAGGCACTACGGTGCTGAAGCAACAAATGCCCTGCTTCCAGGAAAAGCCTCTAAGCATCAGGTAACATCAAATCGTACCCCAAACCGACACAGGTGGTCAGGTAGAGAATACCAAGGCGCTTGAGAGAACTCGGGTGAAGGAACTAGGCAAAATGGTGCCGTAACTTCGGGAGAAGGCACGCTGATATGTAGGTGAAGCGACTTGCTCGTGGAGCTGAAATCAGTCGAAGATACCAGCTGGCTGCAACTGTTTATTAAAAACACAGCACTGTGCAAACACGAAAGTGGACGTATACGGTGTGACGCCTGCCCGGTGCCGGAAGGTTAATTGATGGGGTTAGCGCAAGCGAAGCTCTTGATCGAAGCCCCGGTAAACGGCGGCCGTAACTATAACGGTCCTAAGGTAGCGAAATTCCTTGTCGGGTAAGTTCCGACCTGCACGAATGGCGTAATGATGGCCAGGCTGTCTCCACCCGAGACTCAGTGAAATTGAACTCGCTGTGAAGATGCAGTGTACCCGCGGCAAGACGGAAAGACCCCGTGAACCTTTACTATAGCTTGACACTGAACATTGAGCCTTGATGTGTAGGATAGGTGGGAGGCTTTGAAGTGTGGACGCCAGTCTGCATGGAGCCGACCTTGAAATACCACCCTTTAATGTTTGATGTTCTAACGTTGACCCGTAATCCGGGTTGCGGACAGTGTCTGGTGGGTAGTTTGACTGGGGCGGTCTCCTCCTAAAGAGTAACGGAGGAGCACGAAGGTTGGCTAATCCTGGTCGGACATCAGGAGGTTAGTGCAATGGCATAAGCCAGCTTGACTGCGAGCGTGACGGCGCGAGCAGGTGCGAAAGCAGGTCATAGTGATCCGGTGGTTCTGAATGGAAGGGCCATCGCTCAACGGATAAAAGGTACTCCGGGGATAACAGGCTGATACCGCCCAAGAGTTCATATCGACGGCGGTGTTTGGCACCTCGATGTCGGCTCATCACATCCTGGGGCTGAAGTAGGTCCCAAGGGTATGGCTGTTCGCCATTTAAAGTGGTACGCGAGCTGGGTTTAGAACGTCGTGAGACAGTTCGGTCCCTATCTGCCGTGGGCGCTGGAGAACTGAGGGGGGCTGCTCCTAGTACGAGAGGACCGGAGTGGACGCATCACTGGTGTTCGGGTTGTCATGCCAATGGCACTGCCCGGTAGCTAAATGCGGAAGAGATAAGTGCTGAAAGCATCTAAGCACGAAACTTGCCCCGAGATGAGTTCTCCCTGACTCCTTGAGGGTCCTGAAGGAACGTTGAAGACGACGACGTTGATAGGCCGGGTGTGTAAGCGCAGCGATGCGTTGAGCTAACCGGTACTAATGAACCGTGAGGCTTAACCTT"}, "NCBI_taxonomy": {"NCBI_taxonomy_name": "Escherichia coli CFT073", "NCBI_taxonomy_id": "199310", "NCBI_taxonomy_cvterm_id": "36763"}, "protein_sequence": {"accession": "", "sequence": ""}}}}, "ARO_category": {"36454": {"category_aro_name": "determinant of macrolide resistance", "category_aro_cvterm_id": "36454", "category_aro_accession": "3000315", "category_aro_description": "Enzymes, other proteins or other gene products shown clinically to confer resistance to macrolide antibiotics."}, "35950": {"category_aro_name": "antibiotic resistant gene variant or mutant", "category_aro_cvterm_id": "35950", "category_aro_accession": "0000031", "category_aro_description": "Resistance to antibiotics is often conferred by single nucleotide polymorphisms (SNPs) and other mutations in target genes."}}, "ARO_name": "Escherichia coli 23S rRNA with mutation conferring resistance to erythromycin and telithromycin", "model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "ARO_id": "41258", "model_name": "Escherichia coli 23S rRNA with mutation conferring resistance to macrolide antibiotics", "model_type_id": "40295"}, "2825": {"model_id": "2825", "ARO_accession": "3004180", "model_param": {"blastn_bit_score": {"param_value": "5300", "param_type_id": "41093", "param_type": "BLASTN bit-score", "param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}, "snp": {"param_type": "single resistance variant", "param_value": {"7947": "A2088C", "7948": "C2471U"}, "clinical": {"7947": "A2088C", "7948": "C2471U"}, "param_type_id": "36301", "param_description": "A nucleotide or amino acid substitution that confers elevated resistance to antibiotic(s) relative to wild type. The most common type encoded in the CARD is an amino acid substitution gleaned from the literature with format [wild-type][position][mutation], e.g. R184Q. When present in the associated gene or protein, a single resistance variant confers resistance to an antibiotic drug or drug class. Single resistance variants are used by the protein variant and rRNA mutation models to detect antibiotic resistance from submitted sequences."}}, "ARO_description": "Point mutation in the 23S rRNA of Halobacterium halobium shown to confer resistance to chloramphenicol", "model_sequences": {"sequence": {"4170": {"dna_sequence": {"fmax": "5815", "fmin": "2910", "accession": "X03407.1", "strand": "+", "sequence": "GTGGCTACTGTGCCACCTGGTGGATAGCTCGGCTCGGATGCCGACGAAGGACGTGCCAAGCTGCGATAAGCCTGAGGGAGCCGCACGGAGGCTAAGAACTCAGGATCTCCTAATGGGAATCCCTATAACAATTGCCTTGCGCAATGGGGAACGGCCGGAATTGAAGCATCTCAGTACGGCCAGGAAGAGAAATCGAATGAGACGCCGTTAGTAATGGCGAATGAACGCGGCACAGTCCAAACCGAAGCCTTCGGGCAATGTGGTGTTCGGACTGACTTTCATCGTTTGACCGTTCGTGTGAAGTCTCCTGAAACGGAGCGCGATACAGGGTGACAGCCCCGTATCACGGACCAGTACGACGTGCGTCAGCTCCAGAGTAGCGGGGGTTGGAAATCCCTCGTGAATTGTGGCAGGCATCGACTGCCAAGACTAAGTACTCTCCGAGACCGATAGTGAACAAGTAGTGTGAACGAACGCTGAAAAGCACCCCACAAAGGGGGGTGAAATAGGGCTTGAAATCAGGTGGCGATGGAGCGACGGGGCATAAAAGGCCTCTCTGGGAACGACTTGAGTGCAAACTCATGGTAGGACCTGAGAGGAGCCGATGTTCCGTCGTACGTTTTGAAAAACGAGCTAGGGAGTGTGCCTGTTTGACGAGTCTAACCGGAGTATCCGGGAAGGCGTAGGGAAACCAATATGGCCGCGGCATTGCGAGGGCCACCGTGTTCAAGCGCGGGGAGTCAAACGGGCACGACCCGAAACCCGGTGATCTACGCGTGGGCAAGGTGAAGCATGGCGAAAGCCATGTGGAGGCCTGTTAGGGTTGGTGTCTTTCAACACCCTCCCGTGACCTACGTGTAGGGGTGAAAGGCCCATCGAACCGGGCAACAGCTGGTTCCAACCGAAACATGTCGAAGCATGACCTCTGCCGAGGTAGTTCGTGGGGTAGAGCGACCGATTGGGGAGTTCAACTCCGAGAGGAGTTGTCTCCCCTGTCAAACTCCAAACCTACGGACGCCGTCGACGCAGGGAATCCGGTGTGCGGGGTAAGCCTGTGCACCGTGAGGGAGACAACCCAGAGTTAGGTTAAGGTCCCAAAGTGCGAGCTAAGTGCGATTGAAGGTGGTCTCGAGCCCTAGACAGCCGGGAGGTGAGCTTAGAAGCAGCTACCCTCTAAGAAAAGCGTAACAGCTTACCGGCCGAGGTTCGAGGCGCCCAAAATGATCGGGGCTTAAGTTCGCCACCGAGACCTAACGGCACGGGTAACACCGTGATCCAGTAGGTTGGCATTCTGTTCGGGTGGAAGCTCGGGTGAGAACTCGAGTGGACCGAGTGGAAAAGAAAATCCTGGCCATAGTAGCAGCGTTAGTCGGGTAAGAATCCCGATGGCCGAAAGAGCAAGGGTTCCTCGGCAATGCTTATCAGCCGAGGGTTAGCCGATCCTAAGGCCCGTCGTAATTCGAGCGGGTCAAAAGGGAAACTGGTTAATATTCCAGTGCCACCGTACATTGAAAGTCGACGCCTCGGAGCAGCTTGAGCCGGGCATTCGCCCGGTCGAACCGTCGAAGTTCGTGGAAGCCGTAATGGCAGGAAGCGAACGAACGTCGGAACAGGGAAACTCAAGTCAATCTGGGGCCCGTGAAAAGGCGAGTACGGTGTTCGTACCGAGATCCGACACAGGTGCTCTGGCAGAGGAAGCCAAGGCCTGTCGGGAATAACCGACGTTAGGGAATTCGGCAAGTTAGTCCCGTAAGTTCGCGATAAGGGATGCCTGCCACGCAATGAGGCAGGTCGCAGTGACTCGGAGGCTCCGACTGTCTAATAACAACATAGGTGACCGCAAATCCGCAAGGACGCGTACGGTCACTGAATCCTGCCCAGTGCGGGTATCTGAACACCCAGTACAATGGGGCGAAGGACCCGTTAACGGCGGGGGTAACTATGACCCTCTTAAGGTAGCGTAGTACCTTGCCGCTTCAGTAGCGGCTTGCATGAATGGATCAACGAGAGCCTCACTGTCCCAACGTTGGGCCCGGTGAACTGTACGTTCCAGTGCGGAGTCTGGAGACCCCCAAGGGGAAGCGAAGACCCTATAGAGCTTTACTGCAGGCTGTCGCTGGGACACGGTCGCTGATGTGCAGAGTAGGTAGGAGACGTTACACAGGTACGTGCGCTAGCACGCCACCGAGTCACACATGAAACACTACCCGTCAGTGACTGTGACCCTCACTCCGGGAGGAGGACACCGGTAGCCGGGCAGTTTGACTGGGGCGGTACGCGCTTGAAAAGATATCGAGCGCGCCCTAAGCCTATCTCAGCCGAGTCAGAGACTCGGCGAAGAGTGCAAGAGCATAAGATAGGCTGACAGTGTCCTACACAACGAGGGACGCTGACGCGAAAGCTGGTCTAGCGAACCAATTAGGCTGCTTGATGCGGCCAATTGCTGACAGAAAAGCTACCTTAGGGATAACAGAGTCGTCACTCGCAAGAGCACATATCGACCGAGTGGCTGCTACCTCGATGTCGGTTCCCTCCATCCTGCCCGTGCAGAAGCGGGCAAGGGTGAGGTTGTTCGCCTATTAAAGGAGGTCGTGAGCTGGGTTTAGACCGTCGTGAGACAGGTCGGCTGCTATCTATTGGGGGTGTTTTGGTGCTTGACAGGAACGTTCGTATAGTACGAGAGGAACTACGAACGGGTGCCACTGGTGTATCGGTTGTCCGAGAGGGCATGTGCCGAGCAGCTACGCACCACGGGGTAAGAGCTGAATGCATCTAAGCTCGAAACCCACCTGGAAAAGAAGCACCACTGAGACCGCTCGTAGAAGACGAGTTCGATAGACTTGGGGTGTACGCGTCGAGGCAACGAGACGTTTAGCCCGCGAGTACTAACAGGTCAATGCCACAC"}, "NCBI_taxonomy": {"NCBI_taxonomy_name": "Halobacterium salinarum", "NCBI_taxonomy_id": "2242", "NCBI_taxonomy_cvterm_id": "40807"}, "protein_sequence": {"accession": "", "sequence": ""}}}}, "ARO_category": {"36191": {"category_aro_name": "determinant of phenicol resistance", "category_aro_cvterm_id": "36191", "category_aro_accession": "3000052", "category_aro_description": "Enzymes, other proteins or other gene products shown clinically to confer resistance to phenicol (chloramphenicol) antibiotics. These include chloramphenicol acetyltransferase (CAT) enzymes, which are found in a large number of species."}, "35950": {"category_aro_name": "antibiotic resistant gene variant or mutant", "category_aro_cvterm_id": "35950", "category_aro_accession": "0000031", "category_aro_description": "Resistance to antibiotics is often conferred by single nucleotide polymorphisms (SNPs) and other mutations in target genes."}}, "ARO_name": "Halobacterium halobium 23S rRNA mutation conferring resistance to chloramphenicol", "model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "ARO_id": "41333", "model_name": "Halobacterium halobium 23S rRNA mutation conferring resistance to chloramphenicol", "model_type_id": "40295"}, "2826": {"model_id": "2826", "ARO_accession": "3004181", "model_param": {"blastn_bit_score": {"param_value": "5000", "param_type_id": "41093", "param_type": "BLASTN bit-score", "param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}, "snp": {"param_type": "single resistance variant", "param_value": {"7949": "A2129C"}, "clinical": {"7949": "A2129C"}, "param_type_id": "36301", "param_description": "A nucleotide or amino acid substitution that confers elevated resistance to antibiotic(s) relative to wild type. The most common type encoded in the CARD is an amino acid substitution gleaned from the literature with format [wild-type][position][mutation], e.g. R184Q. When present in the associated gene or protein, a single resistance variant confers resistance to an antibiotic drug or drug class. Single resistance variants are used by the protein variant and rRNA mutation models to detect antibiotic resistance from submitted sequences."}}, "ARO_description": "Point mutation in the 23S rRNA of Streptococcus pneumoniae shown to confer resistance to macrolides and streptogramins types of antibiotics", "model_sequences": {"sequence": {"4171": {"dna_sequence": {"fmax": "1695312", "fmin": "1692408", "accession": "NZ_CP018138.1", "strand": "-", "sequence": "TTTGGATAAGTCCTCGAGCTATTAGTATTAGTCCGCTACATGTGTCGCCACACTTCCACTTCTAACCTATCTACCTGATCATCTCTCAGGGCTCTTACTGATATATAATCATGGGAAATCTCATCTTGAGGTGGGTTTCACACTTAGATGCTTTCAGCGTTTATCCCTTCCCTACATAGCTACCCAGCGATGCCTTTGGCAAGACAACTGGTACACCAGCGGTAAGTCCACTCTGGTCCTCTCGTACTAGGAGCAGATCCTCTCAAATTTCCTACGCCCGCGACGGATAGGGACCGAACTGTCTCACGACGTTCTGAACCCAGCTCGCGTGCCGCTTTAATGGGCGAACAGCCCAACCCTTGGGACCGACTACAGCCCCAGGATGCGACGAGCCGACATCGAGGTGCCAAACCTCCCCGTCGATGTGAACTCTTGGGGGAGATAAGCCTGTTATCCCCAGGGTAGCTTTTATCCGTTGAGCGATGGCCCTTCCATACGGAACCACCGGATCACTAAGCCCGACTTTCGTCCCTGCTCGAGTTGTAGCTCTCGCAGTCAAGCTCCCTTATACCTTTATACTCTGCGAATGATTTCCAACCATTCTGAGGGAACCTTTGGGCGCCTCCGTTACCTTTTAGGAGGCGACCGCCCCAGTCAAACTGCCCGTCAGACACTGTCTCCGATAGGGATCACCTATCTGGGTTAGAGTGGCCATAACACAAGGGTAGTATCCCAACAGCGTCTCCTTCGAAACTGGCGTCCCGATCTCTTAGACTCCTACCTATCCTGTACATGTGGTACAGACACTCAATATCAAACTGCAGTAAAGCTCCATGGGGTCTTTCCGTCCTGTCGCGGGTAACCTGCATCTTCACAGGTACTAAAATTTCACCGAGTCTCTCGTTGAGACAGTGCCCAAATCATTACGCCTTTCGTGCGGGTCGGAACTTACCCGACAAGGAATTTCGCTACCTTAGGACCGTTATAGTTACGGCCGCCGTTTACTGGGGCTTCAATTCATACCTTCGCTTACGCTAAGCACTCCTCTTAACCTTCCAGCACCGGGCAGGCGTCACCCCCTATACATCATCTTACGATTTAGCAGAGAGCTGTGTTTTTGATAAACAGTTGCTTGGGCCTATTCACTGCGGCTGACCTAAAGTCAGCACCCCTTCTCCCGAAGTTACGGGGTCATTTTGCCGAGTTCCTTAACGAGAGTTCTCTCGCTCACCTGAGGCTACTCGCCTCGACTACCTGTGTCGGTTTGCGGTACGGGTAGAGTATGTTTAAACGCTAGAAGCTTTTCTTGGCAGTGTGACGTCACTAACTTCGCTACTAAACTTCGCTCCCCATCACAGCTCAATGTTATAGAATTAAGCATTTGACTCAATTCACACCTCACTGCTTAGACAGACTCTTCCAATCGTCTGCTTTAGTTAGCCTACTGCGTCCCTCCATCACTACATACTCTAGTACAGGAATATCAACCTGTTGTCCATCGGATACACCTTTCGGTCTCTCCTTAGGTCCCGACTAACCCAGGGCGGACGAGCCTTCCCCTGGAAACCTTAGTCTTACGGTGGACAGGATTCTCACCTGTCTTTCGCTACTCATACCGGCATTCTCACTTCTATGCGTTCCAGCACTCCTCACGGTATACCTTCATCACACATAGAACGCTCTCCTACCATACCTATAAAGGTATCCACAGCTTCGGTAAATTGTTTTAGCCCCGGTACATTTTCGGCGCAGGGTCACTCGACTAGTGAGCTATTACGCACTCTTTGAATGAATAGCTGCTTCTAAGCTAACATCCTAGTTGTCTGTGCAACCCCACATCCTTTTCCACTTAACAATTATTTTGGGACCTTAGCTGGTGGTCTGGGCTGTTTCCCTTTCGACTACGGATCTTAGCACTCGCAGTCTGACTGCCGACCATAATTCATTGGCATTCGGAGTTTATCTGAGATTGGTAATCCGGGATGGACCCCTCACCCAAACAGTGCTCTACCTCCAAGAATCTCTAATGTCGACGCTAGCCCTAAAGCTATTTCGGAGAGAACCAGCTATCTCCAAGTTCGTTTGGAATTTCTCCGCTACCCACAAGTCATCCAAGCACTTTTCAACGTGCCCTGGTTCGGTCCTCCAGTGCGTCTTACCGCACCTTCAACCTGCTCATGGGTAGGTCACATGGTTTCGGGTCTACGTCATGATACTAATTCGCCCTGTTCAGACTCGGTTTCCCTACGGCTCCGTCTCTTCAACTTAACCTCGCATCATAACGTAACTCGCCGGTTCATTCTACAAAAGGCACGCTCTCACCCATTAACGGGCTCGAACTTGTTGTAGGCACACGGTTTCAGGTTCTATTTCACTCCCCTCCCGGGGTGCTTTTCACCTTTCCCTCACGGTACTGGTTCACTATCGGTCACTAGGGAGTATTTAGGGTTGGGAGATGGTCCTCCCAGATTCCGACGGGATTTCACGTGTCCCGCCGTACTCAGGATACTGCTAGGTACAAAGACTATTTTAAATACGAGGCTATTACTCTCTTTGGCTGATCTTCCCAAATCATTCTTCTATAATCTTTGAGTCCACATTGCAGTCCTACAACCCCGAAGAGTAAACTCTTCGGTTTGCCCTTCTGCCGTTTCGCTCGCCGCTACTAAGGCAATCGCTTTTGCTTTCTCTTCCTGCAGCTACTTAGATGTTTCAGTTCACTGCGTCTTCCTCCTCACATCCTTAACAGATGTGGGTAACAGGTATTACCTGTTGGGTTCCCCCATTCGGAAATCCCTGGATCATCGCTTACTTACAGCTACCCAAGGTATATCGTCGTTTGTCACGTCCTTCGTCGGCTCCTAGTGCCAAGGCATCCACCGTGCGCCCTTATTAACTTAACCT"}, "NCBI_taxonomy": {"NCBI_taxonomy_name": "Streptococcus pneumoniae", "NCBI_taxonomy_id": "1313", "NCBI_taxonomy_cvterm_id": "35917"}, "protein_sequence": {"accession": "", "sequence": ""}}}}, "ARO_category": {"35950": {"category_aro_name": "antibiotic resistant gene variant or mutant", "category_aro_cvterm_id": "35950", "category_aro_accession": "0000031", "category_aro_description": "Resistance to antibiotics is often conferred by single nucleotide polymorphisms (SNPs) and other mutations in target genes."}, "36454": {"category_aro_name": "determinant of macrolide resistance", "category_aro_cvterm_id": "36454", "category_aro_accession": "3000315", "category_aro_description": "Enzymes, other proteins or other gene products shown clinically to confer resistance to macrolide antibiotics."}, "36379": {"category_aro_name": "determinant of streptogramin resistance", "category_aro_cvterm_id": "36379", "category_aro_accession": "3000240", "category_aro_description": "Ezymes, other proteins or other gene products shown clinically to confer resistance to streptogramin antibiotics."}}, "ARO_name": "Streptococcus pneumoniae 23S rRNA mutation conferring resistance to macrolides and streptogramins antibiotics", "model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "ARO_id": "41334", "model_name": "Streptococcus pneumoniae 23S rRNA mutation conferring resistance to macrolides and streptogramins antibiotics", "model_type_id": "40295"}, "2827": {"model_id": "2827", "ARO_accession": "3004184", "model_param": {"41339": {"param_value": {"8013": "ntG-11A"}, "param_type_id": "41339", "param_type": "mutation in regulatory element", "param_description": "An AMR detection model parameter to describe mutations in regulatory elements upstream of a coding DNA region shown to confer resistance to an antibiotic drug or drug class. Regulatory element mutations in CARD use the following notation: nt[wild-type][-][position][mutation]."}, "blastp_bit_score": {"param_value": "300", "param_type_id": "40725", "param_type": "BLASTP bit-score", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}, "snp": {"param_type": "single resistance variant", "param_value": {"7951": "G8R"}, "clinical": {"7951": "G8R"}, "param_type_id": "36301", "param_description": "A nucleotide or amino acid substitution that confers elevated resistance to antibiotic(s) relative to wild type. The most common type encoded in the CARD is an amino acid substitution gleaned from the literature with format [wild-type][position][mutation], e.g. R184Q. When present in the associated gene or protein, a single resistance variant confers resistance to an antibiotic drug or drug class. Single resistance variants are used by the protein variant and rRNA mutation models to detect antibiotic resistance from submitted sequences."}}, "ARO_description": "ribD is a Mycobacterium tuberculosis riboflavin biosynthesis enzyme. Point mutations in ribD cause enzyme overexpression, which allows the C-terminal reductase domain to act as an alternative dihydrofolate reductase. Thus, mutations in ribD confer resistance to DHFR inhibitors such as para-aminosalicylic acid.", "model_sequences": {"sequence": {"4172": {"dna_sequence": {"fmax": "2987615", "fmin": "2986838", "accession": "NC_000962.3", "strand": "+", "sequence": "ATGCCCGACTCTGGTCAGCTCGGAGCCGCTGACACCCCGCTAAGGCTGCTCAGCTCGGTGCATTACCTCACCGACGGCGAACTCCCCCAGCTTTACGACTATCCGGATGACGGCACCTGGTTGCGGGCGAACTTCATCAGCAGCTTGGACGGCGGCGCTACCGTCGATGGCACCAGCGGGGCGATGGCCGGGCCCGGCGACCGATTCGTCTTCAACCTGTTGCGTGAACTTGCCGACGTCATCGTGGTCGGCGTGGGCACCGTGCGCATTGAGGGCTACTCCGGCGTCCGGATGGGTGTCGTCCAGCGCCAGCACCGGCAGGCCCGAGGCCAAAGCGAAGTTCCGCAACTGGCAATCGTCACCAGGTCCGGTCGCCTTGACCGTGACATGGCGGTATTCACCCGGACCGAGATGGCACCGTTGGTGCTCACCACCACGGCGGTCGCCGATGACACGCGCCAGCGGCTCGCGGGCCTCGCCGAGGTGATCGCGTGCTCCGGCGACGATCCGGGCACGGTCGATGAGGCAGTGCTCGTGTCCCAGCTCGCGGCTCGCGGTCTGCGCCGGATCCTTACCGAAGGCGGGCCGACGTTGCTCGGGACATTCGTCGAGCGTGACGTGCTCGACGAGCTGTGTCTGACGATCGCCCCCTACGTCGTCGGCGGCCTGGCGCGCCGCATAGTGACGGGACCCGGGCAGGTGCTGACCCGGATGCGCTGTGCCCATGTCCTCACCGACGACTCCGGCTACCTGTACACCCGCTACGTCAAGACCTGA"}, "NCBI_taxonomy": {"NCBI_taxonomy_name": "Mycobacterium tuberculosis H37Rv", "NCBI_taxonomy_id": "83332", "NCBI_taxonomy_cvterm_id": "39507"}, "protein_sequence": {"accession": "NP_217187.1", "sequence": "MPDSGQLGAADTPLRLLSSVHYLTDGELPQLYDYPDDGTWLRANFISSLDGGATVDGTSGAMAGPGDRFVFNLLRELADVIVVGVGTVRIEGYSGVRMGVVQRQHRQARGQSEVPQLAIVTRSGRLDRDMAVFTRTEMAPLVLTTTAVADDTRQRLAGLAEVIACSGDDPGTVDEAVLVSQLAARGLRRILTEGGPTLLGTFVERDVLDELCLTIAPYVVGGLARRIVTGPGQVLTRMRCAHVLTDDSGYLYTRYVKT"}}}}, "ARO_category": {"36520": {"category_aro_name": "antibiotic target replacement protein", "category_aro_cvterm_id": "36520", "category_aro_accession": "3000381", "category_aro_description": "Alternate proteins that have the same functions as other antibiotic target proteins, but are structurally different and thus resistant to antibiotics. These can replace the activity of other antibiotic-sensitive proteins in the presence of antibiotics."}, "41298": {"category_aro_name": "determinant of aminosalicylate resistance", "category_aro_cvterm_id": "41298", "category_aro_accession": "3004151", "category_aro_description": "Enzymes, other proteins or other gene products shown clinically to confer resistance to aminosalicylates, esp. para-aminosalicylic acid, commonly used to treat Mycobacterium infections."}}, "ARO_name": "Mycobacterium tuberculosis ribD with mutation conferring resistance to para-aminosalicylic acid", "model_type": "protein variant model", "model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "ARO_id": "41337", "model_name": "Mycobacterium tuberculosis ribD with mutation conferring resistance to para-aminosalicylic acid", "model_type_id": "40293"}, "2820": {"model_id": "2820", "ARO_accession": "3004174", "model_param": {"blastn_bit_score": {"param_value": "5000", "param_type_id": "41093", "param_type": "BLASTN bit-score", "param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}, "snp": {"param_type": "single resistance variant", "param_value": {"7939": "A2663C", "7938": "A2109C"}, "clinical": {"7939": "A2663C", "7938": "A2109C"}, "param_type_id": "36301", "param_description": "A nucleotide or amino acid substitution that confers elevated resistance to antibiotic(s) relative to wild type. The most common type encoded in the CARD is an amino acid substitution gleaned from the literature with format [wild-type][position][mutation], e.g. R184Q. When present in the associated gene or protein, a single resistance variant confers resistance to an antibiotic drug or drug class. Single resistance variants are used by the protein variant and rRNA mutation models to detect antibiotic resistance from submitted sequences."}}, "ARO_description": "Point mutation in the 23S rRNA of Chlamydia trachomatis shown to confer resistance to macrolide type antibiotics", "model_sequences": {"sequence": {"4165": {"dna_sequence": {"fmax": "2940", "fmin": "0", "accession": "NR_076160.1", "strand": "+", "sequence": "CAGACCAAGTTAATAAGAGCTATTGGTGGATGCCTTGGCATTGACAGGCGAAGAAGGACGCGAATACCTGCGAAAAGCTCCGGCGAGCTGGTGATAAGCAAAGACCCGGAGGTATCCGAATGGGGAAACCCGGTAGAGTAATAGACTACCATTGCATGCTGAATACATAGGTATGCAAAGCGACACCTGCCGAACTGAAACATCTTAGTAAGCAGAGGAAAAGAAATCGAAGAGATTCCCTGTGTAGCGGCGAGCGAAAGGGGAATAGCCTAAACCGAGCTGATAAGGCTCGGGGTTGTAGGATTGAGGATAAAGGATCAGGACTCCTAGTTGAACACATCTGGAAAGATGGATGATACAGGGTGATAGTCCCGTAGACGAAAGGAGAGAAAGACCGACCTCAACACCTGAGTAGGACTAGACACGTGAAACCTAGTCTGAATCTGGGGAGACCACTCTCCAAGGCTAAATACTAGTCAATGACCGATAGTGAACCAGTACTGTGAAGGAAAGGCGAAAAGAACCCTTGTTAAGGGAGTGAAATAGAACCTGAAACCAGTAGCTTACAAGCGGTCGGAGACCAATGGCCCGTAAGGGTCAAGGTTGACGGCGTGCCTTTTGCATGATGAGCCAGGGAGTTAAGCTAAACGGCGAGGTTAAGGGATATACATTCCGGAGCCGGAGCGAAAGCGAGTTTTAAAAGAGCGAAGAGTCGTTTGGTTTAGACACGAAACCAAGTGAGCTATTTATGACCAGGTTGAAGCATGGGTAAAACTATGTGGAGGACCGAACTAGTACCTGTTGAAAAAGGTTTGGATGAGTTGTGAATAGGGGTGAAAGGCCAATCAAACTTGGAGATATCTTGTTCTCTCCGAAATAACTTTAGGGTTAGCCTCGGATAATAAGCTTTTGGGGGTAGAGCACTGAATTCTAGCGGGGGCCTACCGGCTTACCAACGGAAATCAAACTCCGAATACCAGAAGCGAGTCCGGGAGATAGACAGCGGGGGCTAAGCTTCGTTGTCGAGAGGGGAACAGCCCAGACCGCCGATTAAGGTCCCTAATTTTATGCTAAGTGGGTAAGGAAGTGATGATTCGAAGACAGTTGGAATGTTGGCTTAGAGGCAGCAATCATTTAAAGAGTGCGTAACAGCTCACCAATCGAGAATCATTGCGCCGATAATAAACGGGACTAAGCATAAAACCGACATCGCGGGTGTGTCGATAAGACACGCGGTAGGAGAGCGTAGTATTCAGCAGAGAAGGTGTACCGGAAGGAGCGCTGGAGCGGATACTAGTGAAGATCCATGGCATAAGTAACGATAAAGGGAGTGAAAATCTCCCTCGCCGTAAGCCCAAGGTTTCCAGGGTCAAGCTCGTCTTCCCTGGGTTAGTCGGCCCCTAAGTTGAGGCGTAACTGCGTAGACGATGGAGCAGCAGGTTAAATATTCCTGCACCACCTAAAACTATAGCGAAGGAATGACGGAGTAAGTTAAGCACGCGGACGATTGGAAGAGTCCGTAGAGCGATGAGAACGGTTAGTAGGCAAATCCGCTAACATAAGATCAGGTCGCGATCAAGGGGAATCTTCGGGGGAACCGATGGTGTGGAGCGAGGCTTTCAAGAAATAATTTCTAGCTGTTGATGGTGACCGTACCAAAACCGACACAGGTGGGCGAGATGAATATTCTAAGGCGCGCGAGATAACTTTCGTTAAGGAACTCGGCAAATTATCCCCGTAACTTCGGAATAAGGGGAGCCTTTTAGGGTGACTATGGAACGATAGGAGCCCCGGGGGGCCGCAGAGAAATGGCCCAGGCGACTGTTTAGCAAAAACACAGCACTATGCAAACCTCTAAGGGGAAGTATATGGTGTGACGCCTGCCCAATGCCAAAAGGTTAAAGGGATATGTCAGCTGCAAAGCGAAGCATTGAACCTAAGCCCTGGTGAATGGCCGCCGTAACTATAACGGTGCTAAGGTAGCGAAATTCCTTGTCGGGTAAGTTCCGACCTGCACGAATGGTGTAACGATCTGGGCACTGTCTCAACGAAAGACTCGGTGAAATTGTAGTAGCAGTGAAGATGCTGTTTACCCGCGAAAGGACGAAAAGACCCCGTGAACCTTTACTGTACTTTGGTATTGATTTTTGGTTTGTTATGTGTAGGATAGCCAGGAGACTAAGAACACTCTTCTTCAGGAGAGTGGGAGTCAACGTTGAAATACTGGTCTTAACAAGCTGGGAATCTAACATTATTCCATGAATCTGGAAGATGGACATTGCCAGACGGGCAGTTTTACTGGGGCGGTATCCTCCTAAAAAGTAACGGAGGAGCCCAAAGCTTATTTCATCGTGGTTGGCAATCACGAGTAGAGCGTAAAGGTATAAGATAGGTTGACTGCAAGACCAACAAGTCGAGCAGAGACGAAAGTCGGGCTTAGTGATCCGGCGGTGGAAAGTGGAATCGCCGTCGCTTAACGGATAAAAGGTACTCCGGGGATAACAGGCTGATCGCCACCAAGAGTTCATATCGACGTGGCGGTTTGGCACCTCGATGTCGGCTCATCGCATCCTGGGGCTGGAGAAGGTCCCAAGGGTTTGGCTGTTCGCCAATTAAAGCGGTACGCGAGCTGGGTTCAAAACGTCGTGAGACAGTTTGGTCTCTATCCTTCGTGGGCGCAGGATACTTGAGAGGAGCTGTTCCTAGTACGAGAGGACCGGAATGGACGAACCAATGGTGTGTCGGTTGTTTTGCCAAGGGCATAGCCGGGTGGCTCACGTTCGGAAAGGATAAGCATTGAAAGCATCTAAATGCCAAGCCTCCCTCAAGATAAGGTATCCCAATGAGACTCCATGTAGACTACGTGGTTGATAGGTTGGAGGTGTAAGCACAGTAATGTGTTCAGCTAACCAATACTAATAAGTCCAAAGACTTGGTCTT"}, "NCBI_taxonomy": {"NCBI_taxonomy_name": "Chlamydia trachomatis", "NCBI_taxonomy_id": "813", "NCBI_taxonomy_cvterm_id": "40470"}, "protein_sequence": {"accession": "", "sequence": ""}}}}, "ARO_category": {"35950": {"category_aro_name": "antibiotic resistant gene variant or mutant", "category_aro_cvterm_id": "35950", "category_aro_accession": "0000031", "category_aro_description": "Resistance to antibiotics is often conferred by single nucleotide polymorphisms (SNPs) and other mutations in target genes."}, "36454": {"category_aro_name": "determinant of macrolide resistance", "category_aro_cvterm_id": "36454", "category_aro_accession": "3000315", "category_aro_description": "Enzymes, other proteins or other gene products shown clinically to confer resistance to macrolide antibiotics."}}, "ARO_name": "Chlamydia trachomatis 23S rRNA with mutation conferring resistance to macrolide antibiotics", "model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "ARO_id": "41325", "model_name": "Chlamydia trachomatis 23S rRNA with mutation conferring resistance to macrolide antibiotics", "model_type_id": "40295"}, "2807": {"model_id": "2807", "ARO_accession": "3004160", "model_param": {"blastn_bit_score": {"param_value": "5000", "param_type_id": "41093", "param_type": "BLASTN bit-score", "param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}, "snp": {"param_type": "single resistance variant", "param_value": {"7899": "G2032U"}, "clinical": {"7899": "G2032U"}, "param_type_id": "36301", "param_description": "A nucleotide or amino acid substitution that confers elevated resistance to antibiotic(s) relative to wild type. The most common type encoded in the CARD is an amino acid substitution gleaned from the literature with format [wild-type][position][mutation], e.g. R184Q. When present in the associated gene or protein, a single resistance variant confers resistance to an antibiotic drug or drug class. Single resistance variants are used by the protein variant and rRNA mutation models to detect antibiotic resistance from submitted sequences."}}, "ARO_description": "Point mutation in the 23S rRNA of Escherichia coli shown clinically to confer resistance to clarithromycin, a macrolide antibiotic", "model_sequences": {"sequence": {"4154": {"dna_sequence": {"fmax": "240063", "fmin": "237159", "accession": "AE014075.1", "strand": "+", "sequence": "GGTTAAGCGACTAAGCGTACACGGTGGATGCCCTGGCAGTCAGAGGCGATGAAGGACGTGCTAATCTGCGATAAGCGTCGGTAAGGTGATATGAACCGTTATAACCGGCGATTTCCGAATGGGGAAACCCAGTGTGTTTCGACACACTATCATTAACTGAATCCATAGGTTAATGAGGCGAACCGGGGGAACTGAAACATCTAAGTACCCCGAGGAAAAGAAATCAACCGAGATTCCCCCAGTAGCGGCGAGCGAACGGGGAGGAGCCCAGAGCCTGAATCAGTGTGTGTGTTAGTGGAAGCGTCTGGAAAGGCGCGCGATACAGGGTGACAGCCCCGTACACAAAAATGCACATGCTGTGAGCTCGATGAGTAGGGCGGGACACGTGGTATCCTGTCTGAATATGGGGGGACCATCCTCCAAGGCTAAATACTCCTGACTGACCGATAGTGAACCAGTACCGTGAGGGAAAGGCGAAAAGAACCCCGGCGAGGGGAGTGAAAAAGAACCTGAAACCGTGTACGTACAAGCAGTGGGAGCATGCTTAGGCGTGTGACTGCGTACCTTTTGTATAATGGGTCAGCGACTTATATTCTGTAGCAAGGTTAACCGAATAGGGGAGCCGAAGGGAAACCGAGTCTTAACTGGGCGTTAAGTTGCAGGGTATAGACCCGAAACCCGGTGATCTAGCCATGGGCAGGTTGAAGGTTGGGTAACACTAACTGGAGGACCGAACCGACTAATGTTGAAAAATTAGCGGATGACTTGTGGCTGGGGGTGAAAGGCCAATCAAACCGGGAGATAGCTGGTTCTCCCCGAAAGCTATTTAGGTAGCGCCTCGTGAACTCATCTCCGGGGGTAGAGCACTGTTTCGGCAAGGGGGTCATCCCGACTTACCAACCCGATGCAAACTGCGAATACCGGAGAATGTTATCACGGGAGACACACGGCGGGTGCTAACGTCCGTCGNGAAGAGGGAAACAACCCAGACCGCCAGCTAAGGTCCCAAAGTCATGGTTAAGTGGGAAACGATGTGGGAAGGCCCAGACAGCCAGGATGTTGGCTTAGAAGCAGCCATCATTTAAAGAAAGCGTAATAGCTCACTGGTCGAGTCGGCCTGCGCGGAAGATGTAACGGGGCTAAACCATGCACCGAAGCTGCGGCAGCGACGCTTATGCGTTGTTGGGTAGGGGAGCGTTCTGTAAGCCTGTGAAGGTGTACTGTGAGGTATGCTGGAGGTATCAGAAGTGCGAATGCTGACATAAGTAACGATAAAGCGGGTGAAAAGCCCGCTCGCCGGAAGACCAAGGGTTCCTGTCCAACGTTAATCGGGGCAGGGTGAGTCGACCCCTAAGGCGAGGCCGAAAGGCGTAGTCGATGGGAAACAGGTTAATATTCCTGTACTTGGTGTTACTGCGAAGGGGGGACGGAGAAGGCTATGTTGGCCGGGCGACGGTTGTCCCGGTTTAAGCGTGTAGGCTGGTTTTCCAGGCAAATCCGGAAAATCAAGGCTGAGGCGTGATGACGAGGCACTACGGTGCTGAAGCAACAAATGCCCTGCTTCCAGGAAAAGCCTCTAAGCATCAGGTAACATCAAATCGTACCCCAAACCGACACAGGTGGTCAGGTAGAGAATACCAAGGCGCTTGAGAGAACTCGGGTGAAGGAACTAGGCAAAATGGTGCCGTAACTTCGGGAGAAGGCACGCTGATATGTAGGTGAAGCGACTTGCTCGTGGAGCTGAAATCAGTCGAAGATACCAGCTGGCTGCAACTGTTTATTAAAAACACAGCACTGTGCAAACACGAAAGTGGACGTATACGGTGTGACGCCTGCCCGGTGCCGGAAGGTTAATTGATGGGGTTAGCGCAAGCGAAGCTCTTGATCGAAGCCCCGGTAAACGGCGGCCGTAACTATAACGGTCCTAAGGTAGCGAAATTCCTTGTCGGGTAAGTTCCGACCTGCACGAATGGCGTAATGATGGCCAGGCTGTCTCCACCCGAGACTCAGTGAAATTGAACTCGCTGTGAAGATGCAGTGTACCCGCGGCAAGACGGAAAGACCCCGTGAACCTTTACTATAGCTTGACACTGAACATTGAGCCTTGATGTGTAGGATAGGTGGGAGGCTTTGAAGTGTGGACGCCAGTCTGCATGGAGCCGACCTTGAAATACCACCCTTTAATGTTTGATGTTCTAACGTTGACCCGTAATCCGGGTTGCGGACAGTGTCTGGTGGGTAGTTTGACTGGGGCGGTCTCCTCCTAAAGAGTAACGGAGGAGCACGAAGGTTGGCTAATCCTGGTCGGACATCAGGAGGTTAGTGCAATGGCATAAGCCAGCTTGACTGCGAGCGTGACGGCGCGAGCAGGTGCGAAAGCAGGTCATAGTGATCCGGTGGTTCTGAATGGAAGGGCCATCGCTCAACGGATAAAAGGTACTCCGGGGATAACAGGCTGATACCGCCCAAGAGTTCATATCGACGGCGGTGTTTGGCACCTCGATGTCGGCTCATCACATCCTGGGGCTGAAGTAGGTCCCAAGGGTATGGCTGTTCGCCATTTAAAGTGGTACGCGAGCTGGGTTTAGAACGTCGTGAGACAGTTCGGTCCCTATCTGCCGTGGGCGCTGGAGAACTGAGGGGGGCTGCTCCTAGTACGAGAGGACCGGAGTGGACGCATCACTGGTGTTCGGGTTGTCATGCCAATGGCACTGCCCGGTAGCTAAATGCGGAAGAGATAAGTGCTGAAAGCATCTAAGCACGAAACTTGCCCCGAGATGAGTTCTCCCTGACTCCTTGAGGGTCCTGAAGGAACGTTGAAGACGACGACGTTGATAGGCCGGGTGTGTAAGCGCAGCGATGCGTTGAGCTAACCGGTACTAATGAACCGTGAGGCTTAACCTT"}, "NCBI_taxonomy": {"NCBI_taxonomy_name": "Escherichia coli CFT073", "NCBI_taxonomy_id": "199310", "NCBI_taxonomy_cvterm_id": "36763"}, "protein_sequence": {"accession": "", "sequence": ""}}}}, "ARO_category": {"35950": {"category_aro_name": "antibiotic resistant gene variant or mutant", "category_aro_cvterm_id": "35950", "category_aro_accession": "0000031", "category_aro_description": "Resistance to antibiotics is often conferred by single nucleotide polymorphisms (SNPs) and other mutations in target genes."}, "36454": {"category_aro_name": "determinant of macrolide resistance", "category_aro_cvterm_id": "36454", "category_aro_accession": "3000315", "category_aro_description": "Enzymes, other proteins or other gene products shown clinically to confer resistance to macrolide antibiotics."}}, "ARO_name": "Escherichia coli 23S rRNA with mutation conferring resistance to clarithromycin", "model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "ARO_id": "41307", "model_name": "Escherichia coli 23S rRNA with mutation conferring resistance to clarithromycin", "model_type_id": "40295"}, "2804": {"model_id": "2804", "ARO_accession": "3004157", "model_param": {"blastp_bit_score": {"param_value": "500", "param_type_id": "40725", "param_type": "BLASTP bit-score", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}, "snp": {"param_type": "single resistance variant", "param_value": {"7878": "I43A", "7879": "R49P", "7874": "E40Q", "7875": "E40G", "7876": "I43T", "7877": "I43S", "7873": "T20P", "7881": "L56V", "7880": "R49W", "7883": "S150G", "7882": "R91W", "7885": "E153A", "7884": "S150C", "7887": "A420V", "7886": "E153G"}, "clinical": {"7878": "I43A", "7879": "R49P", "7874": "E40Q", "7875": "E40G", "7876": "I43T", "7877": "I43S", "7873": "T20P", "7881": "L56V", "7880": "R49W", "7883": "S150G", "7882": "R91W", "7885": "E153A", "7884": "S150C", "7887": "A420V", "7886": "E153G"}, "param_type_id": "36301", "param_description": "A nucleotide or amino acid substitution that confers elevated resistance to antibiotic(s) relative to wild type. The most common type encoded in the CARD is an amino acid substitution gleaned from the literature with format [wild-type][position][mutation], e.g. R184Q. When present in the associated gene or protein, a single resistance variant confers resistance to an antibiotic drug or drug class. Single resistance variants are used by the protein variant and rRNA mutation models to detect antibiotic resistance from submitted sequences."}}, "ARO_description": "Point mutations in the dihydrofolate synthetase folC gene shown clinically to confer resistance to p-aminosalicylic acid or other aminosalicylates. Mutations in folC inhibit bioactivation of PAS and thus confer resistance.", "model_sequences": {"sequence": {"4151": {"dna_sequence": {"fmax": "2747598", "fmin": "2746134", "accession": "NC_000962.3", "strand": "-", "sequence": "TCATTGCGGATCACGACCGAACAAGGTCCGAGCCGCCCCTGCGGTGACAACCGAGCCGGTGATGACGATCCCGGTTCTCGAGAATGCGTCCCCGGCCACATCCGGGTCGGCGGCGGCGTCGTCGACCAGTGAGGTGGCAACGTCGATAGCATCGCGCAGGTTCTCGGCGGTGCGCACCCGGTCGGGTCCGAACCGCTCGCCGGCCGCCAGCGCCAGGGCCTCGACATCCAGCGCCCGCGGCGACCCGTTGTGGGTCACGACGACGGAATCGAACACCGGCTCCAGTGCGGCCAGGATGCCGTCCACGTCCTTGTCGCCCAGCACGCTGAGCACCCCGACCAGAAATCGGAAGTCGAACTCATGCGCCAGCGTTTGTGCCAGAGCACTCGCCCCGGCCGGATTGTGCGCGGCGTCGATGAACACCGTGGGTGCGCTGCGCATGCGCTCCAACCGGCCGGGACTGGTGACGGCGGCAAAGCCGGCCCGGACGGCGTCGCCGTCGAGCTGACGCTGCGCACCGGCACCGAAAAAGGCCTCGACGGAAGCGAGGGCGAGCACCGCGTTGTGCGCCTGGTGTTCACCGTGCAGCGGCAAGTAGATGTCGGAGTAAACCCCGCCGAGGCCCTGCAGTTGCAGTACCTGACCGCCGACCGCGATCTGTCGCCGTAGCACCGCGAATTCGGAATCCTCCCGGGCCACCGACGCGTCGGCGCGCACCGATTCGGCCAGCAGCACCTCCATGACCTTCGGGACCTGACGCCCGATGACCGCGACGGTGTCCGGCGAACCGTCGGGGGCCCGAGTGATGATGCCCGCCTTCTCCCCGGCGATCCCGGCGATATCGGCACCGAGATAGTCGACGTGATCAATGCTGATCGGGGTGATGACGGCGACCGGTGCGTTGATCACGTTGGTGGCGTCCCAACGTCCGCCCATGCCCACCTCGACCACTGCCACGTCGACGGGCGCGTCCGCAAAGGCCGCGAACGCCATCGCGGTGAGCACCTCGAACTTGCTCATCGCCGGGCCACCCTTACCCGCAGAAGCCTGCGACTGCTGGTCGATCAGCGCCACCAACGGCTCGATCTCCCGGTAGGTCGCCACATACTGCGCCGGGCTGATCGGCTTGCCGTCGATCGAAATGCGTTCCACCGGTGACTGCAGGTGTGGGCTGGTGGTTCGGCCGGTGCGCCGGTGCAGCGCGGTGACCAGCGCGTCGACCATGCGCGCCACCGAGGTCTTGCCGTTGGTGCCCGCGATATGGATCGACGGATAGCTGCGTTGGGGCGAGCCCAGCAGGTCCATCAACGCGCTGATCCGGGTCAGGCTCGGATCGATGCGGGTCTCCGGCCAGCGTTGGTCGAGTAGATGCTCAACCTGCAGCAGGGACGCGATCTCGTCCGGAGTGGGCACGACGCCGGTGGCCGATCCCGAGTCAGGCGGGCCGGAATTCGTCGAATTCAT"}, "NCBI_taxonomy": {"NCBI_taxonomy_name": "Mycobacterium tuberculosis H37Rv", "NCBI_taxonomy_id": "83332", "NCBI_taxonomy_cvterm_id": "39507"}, "protein_sequence": {"accession": "NP_216963.1", "sequence": "MNSTNSGPPDSGSATGVVPTPDEIASLLQVEHLLDQRWPETRIDPSLTRISALMDLLGSPQRSYPSIHIAGTNGKTSVARMVDALVTALHRRTGRTTSPHLQSPVERISIDGKPISPAQYVATYREIEPLVALIDQQSQASAGKGGPAMSKFEVLTAMAFAAFADAPVDVAVVEVGMGGRWDATNVINAPVAVITPISIDHVDYLGADIAGIAGEKAGIITRAPDGSPDTVAVIGRQVPKVMEVLLAESVRADASVAREDSEFAVLRRQIAVGGQVLQLQGLGGVYSDIYLPLHGEHQAHNAVLALASVEAFFGAGAQRQLDGDAVRAGFAAVTSPGRLERMRSAPTVFIDAAHNPAGASALAQTLAHEFDFRFLVGVLSVLGDKDVDGILAALEPVFDSVVVTHNGSPRALDVEALALAAGERFGPDRVRTAENLRDAIDVATSLVDDAAADPDVAGDAFSRTGIVITGSVVTAGAARTLFGRDPQ"}}}}, "ARO_category": {"35950": {"category_aro_name": "antibiotic resistant gene variant or mutant", "category_aro_cvterm_id": "35950", "category_aro_accession": "0000031", "category_aro_description": "Resistance to antibiotics is often conferred by single nucleotide polymorphisms (SNPs) and other mutations in target genes."}, "41298": {"category_aro_name": "determinant of aminosalicylate resistance", "category_aro_cvterm_id": "41298", "category_aro_accession": "3004151", "category_aro_description": "Enzymes, other proteins or other gene products shown clinically to confer resistance to aminosalicylates, esp. para-aminosalicylic acid, commonly used to treat Mycobacterium infections."}}, "ARO_name": "Mycobacterium tuberculosis folC with mutation conferring resistance to para-aminosalicylic acid", "model_type": "protein variant model", "model_description": "The protein variant model is an AMR detection model. Protein variant models are similar to protein homolog models - they detect the presence of a protein sequence based on its similarity to a curated reference sequence, but secondarily search submitted query sequences for curated sets of mutations shown clinically to confer resistance relative to wild-type. This model includes a protein reference sequence, a curated BLASTP cut-off, and mapped resistance variants. Mapped resistance variants may include any or all of: single resistance variants, insertions, deletions, co-dependent resistance variants, nonsense SNPs, and/or frameshift mutations. Protein variant model matches to reference sequences are categorized on two criteria: \"strict\" and \"loose\". A strict match has a BLASTP bitscore above the curated BLASTP cutoff value and contains at least one detected mutation from amongst the mapped resistance variants; a loose match has a BLASTP bitscore below the curated BLASTP cutoff value but still contains at least one detected mutation from amongst the mapped resistance variants. Regardless of BLASTP bitscore, if a sequence does not contain one of the mapped resistance variants, it is not considered a match and not detected by the protein variant model.", "ARO_id": "41304", "model_name": "Mycobacterium tuberculosis folC with mutation conferring resistance to para-aminosalicylic acid", "model_type_id": "40293"}, "2805": {"model_id": "2805", "ARO_accession": "3004159", "model_param": {"blastn_bit_score": {"param_value": "5000", "param_type_id": "41093", "param_type": "BLASTN bit-score", "param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}, "snp": {"param_type": "single resistance variant", "param_value": {"7889": "A2072G", "7896": "U2518G", "7897": "A2586U", "7894": "C2069A", "7895": "G2461U", "7892": "G2461A", "7893": "A2517U", "7890": "A2073G", "7891": "G2075U"}, "clinical": {"7889": "A2072G", "7896": "U2518G", "7897": "A2586U", "7894": "C2069A", "7895": "G2461U", "7892": "G2461A", "7893": "A2517U", "7890": "A2073G", "7891": "G2075U"}, "param_type_id": "36301", "param_description": "A nucleotide or amino acid substitution that confers elevated resistance to antibiotic(s) relative to wild type. The most common type encoded in the CARD is an amino acid substitution gleaned from the literature with format [wild-type][position][mutation], e.g. R184Q. When present in the associated gene or protein, a single resistance variant confers resistance to an antibiotic drug or drug class. Single resistance variants are used by the protein variant and rRNA mutation models to detect antibiotic resistance from submitted sequences."}}, "ARO_description": "Point mutations in the 23S rRNA (domain V) of Mycoplasma gallisepticum can confer resistance to pleuromutilin antibiotics (tiamulin and valnemulin). The mutants also show cross-resistance to lincomycin, chloramphenicol and florfenicol.", "model_sequences": {"sequence": {"4152": {"dna_sequence": {"fmax": "83423", "fmin": "80510", "accession": "NC_004829.2", "strand": "+", "sequence": "ATACATAAGTTACTAAGGGCTTATGGTGGATGCCTTGGCACTAGAAGGCGATGAAGGACGTGCAAACCTGCGAAATGCTACGGGGAGCTGGTTGGAAGCGATAATCCGTAGATGTCCGAATGGGGGAACCTGATTAATAGTGATATTAATTATTTAGATCTGAATACATAGGGTCTAAAAGCAATACGTTGTGAAGTGAAACATCTCAGTAGCAACAGGAAAAGAAATCGAAAGAGATTCCGTGTGTAGTGGCGAGCGAAAGCGGAACAGGCCAAACCAAGATTTATCTTGGGGTTATAGGACTGCAATGTGGACTTTGAACTGATAGGAGAAGTAGTTGAAAAGCTACGCGATAAAGGGTTATAGCCCCGTATCTTAAATTGGTTTAATACCTAGCAGGATCCTGAGTACATCGAGAAACGTTATCTTGATGGAAGTCGCCCAGACCATTGGGCAAGCCTAAATACTAACTAGTGACCGATAGCGTATAGTACCGTGAGGGAAAGGTGAAAAGAACCCAGGGATGGGAGTGAAATAGATTCTGAAGCCATATGCCTACAACGTGTCAGAGCACATTAATGTGTGATGGCGTGCGTTTTGAAGTATGAGCCGGCGAGTTATGATAGCAAGCAGGTTAACCTTTAGAAGGGAAGCCGAAGCGAAAGCGAGTTTGAATAGAGCGAATTAAAGTGTTTGTTATTATAGACCCGAAACGGGTTGAGCTAGTCATGGGCAGGTTGAAGTTAGAGTAACATCTAATGGAGGACCGAACCGACTTTCGTTGAAACGACAGCGGATGACCTGTGACTAGGGGTGAAATTCCAATCGAAATCCGTGATAGCTGGTTCTCGTCGAAATAGTTTTAAGACTAGCGTAAGATCATGATCAACTGGAGGTAGAGCTACTGAATGTATGATGGCGCCGCCTTGGTGTACTGAATACAATTAAACTCCGAATGCCAATTGATTTATTCTTGCAGTCAGACAGTGGGGGATAAGCTTCATTGTCACAAGGGGAAGAGCCCAGATCATTAAATAAGGTCCCTAAAATATGCTAAGTGGAAAAGGTTGTTAAAATACTTAAACAGCAAGGATGTTGGCTTAGAAGCAGCCATCGTTTAAAGAGTGCGTAACAGCTCACTTGTCGAGTGTTTTTGCGCCGAAGATGTAACGGGGCTAAGCATATTACCGAATTTATGGATTATTATTCGTAAGAATGATAGTGGTAGACGAGCGTTGTATATGGGATGAAGTCAAACCGTGAGGATTGGTGGACTGTATACAAGTGAGAATGCCGGTGTAAGTAACGCTTGAGAGTGAGAATCTCTCAAACCGATTGACTAAGGTTTCCTGGACGAGGGTCGTCCTTCCAGGGTTAGTCTGGACCTAAGGCGAGGCAGAAATACGTAGTCGATGGAAGAACAGGTTAATATTCCTGTACAAACAAATAGCTGATGGAGTGACGGAGAAGGTTAATGCATCCCCATTATCGGATTTGGGGTTAAATAAGAAGTCTTAAGGGTTGGCAAATCCGCCTTTTTTAAGGACAACTTATGAATACGAGTGAACGCTTTGCAAGTAGCGAAGATGCATACATCACGCTTCCAAGAAAAGCTTCTAGGGTTAACTATTTGTTTCCAGTACCGAGAACGAACACACGTGGTCAAGGAGAAGATCCTAAGGTTAGCGAGTGAACTATAGCTAAGGAACTCTGCAAATTCATCCCGTAAGTTCGCAAGAAGGGATGCTCAATGTAACAGTTGAGCCGCAGTGAAGAACGAGGGGGGACTGTTTAACTAAAACACAGCTCTATGCTAAATCGCAAGATGATGTATATGGGGTGACACCTGCCCAGTGCTGGAAGGTTAAAGAAGGGTGTTAGAGCAATCAAAGCTCCCGACTGAAGCCCCAGTGAACGGCGGCCGTAACTATAACGGTCCTAAGGTAGCGAAATTCCTTGTCGGGTAAATTCCGTCCCGCTTGAATGGTGTAACCATCTCTTGACTGTCTCGGCTATAGACTCGGTGAAATCCAGGTACGGGTGAAGACACCCGTTAGGCGCAACGGGACGGAAAGACCCCATGAAGCTTTACTGTAACTTAATATTGGGCAGAGTTTAGACATATAGAGAATAGGTGGGAGACTTTGAAGCAACTTCGCTAGGAGTTGTGGAGTCACCAGTGGAATACCACCTTTGTTTAAATTCTTCTCTAACTAGTTGCTGTTATCCAGCAATAGGACAGTGTTAGGCGGGCAGTTTGACTGGGGCGGTCGCCTCCCAAAAGGTAACGGAGGCGTGCAAAGGTACCCTCAGCACGGTTGGAAATCGTGTTAAGAGTGTAATGGTATAAGGGTGCTTGACTGTGAGACTAACAGGTCGAACAGGTAAGAAATTAGGTCATAGTGATCCGGTGGTTCAGTATGGAATGGCCATCGCTCAACGGATAAAAGCTACTCTGGGGATAACAGGCTGATACTGCCCAAGAGTTCACATCGACGGCAGTGTTTGGCACCTCGATGTCGACTCATCTCATCCTCGAGCTGAAGCAGGTTCGAAGGGTTCGGCTGTTCGCCGATTAAAGAGATACGTGAGTTGGGTTCAAACCGTCGCGAGACAGGTTGGTCCCTATCTATTGTGCCCGCAGGAAGATTGAAAAGATTTACTCTTAGTACGAGAGGACCGGAGTGAAGACACCTCTTGTGCTCCAGTTGTAGTGCCAACTGCACCGCTGGGTAGCAACGTGTCGAACGGATAAACGCTGAAAGCATCTAAGTGTGAAACCGACTTTAAGAATAATCTTCCCTTCCAGCAATGGAGTAAGAATCGTTGTAGACTACGACGTTGATAGGCTAAAGGTGTAAGTGCCGCGAGGTATTTAGCTGATTAGTACTAATAATTCGAGGACTTAGATTT"}, "NCBI_taxonomy": {"NCBI_taxonomy_name": "Mycoplasma gallisepticum str. R(low)", "NCBI_taxonomy_id": "710127", "NCBI_taxonomy_cvterm_id": "41282"}, "protein_sequence": {"accession": "", "sequence": ""}}}}, "ARO_category": {"35950": {"category_aro_name": "antibiotic resistant gene variant or mutant", "category_aro_cvterm_id": "35950", "category_aro_accession": "0000031", "category_aro_description": "Resistance to antibiotics is often conferred by single nucleotide polymorphisms (SNPs) and other mutations in target genes."}, "40410": {"category_aro_name": "determinant of pleuromutilin resistance", "category_aro_cvterm_id": "40410", "category_aro_accession": "3003753", "category_aro_description": "Enzymes, other proteins or other gene products shown clinically to confer resistance to pleuromutilin antibiotics."}}, "ARO_name": "Mycoplasma gallisepticum 23S rRNA mutation conferring resistance to pleuromutilin antibiotics", "model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "ARO_id": "41306", "model_name": "Mycoplasma gallisepticum 23S rRNA mutation conferring resistance to pleuromutilin antibiotics", "model_type_id": "40295"}, "2806": {"model_id": "2806", "ARO_accession": "3004149", "model_param": {"blastn_bit_score": {"param_value": "5000", "param_type_id": "41093", "param_type": "BLASTN bit-score", "param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}, "snp": {"param_type": "single resistance variant", "param_value": {"7898": "G2032A"}, "clinical": {"7898": "G2032A"}, "param_type_id": "36301", "param_description": "A nucleotide or amino acid substitution that confers elevated resistance to antibiotic(s) relative to wild type. The most common type encoded in the CARD is an amino acid substitution gleaned from the literature with format [wild-type][position][mutation], e.g. R184Q. When present in the associated gene or protein, a single resistance variant confers resistance to an antibiotic drug or drug class. Single resistance variants are used by the protein variant and rRNA mutation models to detect antibiotic resistance from submitted sequences."}}, "ARO_description": "Point mutation in the 23S rRNA of Escherichia coli shown clinically to confer resistance to clindamycin, a lincosamide antibiotic", "model_sequences": {"sequence": {"4153": {"dna_sequence": {"fmax": "240063", "fmin": "237159", "accession": "AE014075.1", "strand": "+", "sequence": "GGTTAAGCGACTAAGCGTACACGGTGGATGCCCTGGCAGTCAGAGGCGATGAAGGACGTGCTAATCTGCGATAAGCGTCGGTAAGGTGATATGAACCGTTATAACCGGCGATTTCCGAATGGGGAAACCCAGTGTGTTTCGACACACTATCATTAACTGAATCCATAGGTTAATGAGGCGAACCGGGGGAACTGAAACATCTAAGTACCCCGAGGAAAAGAAATCAACCGAGATTCCCCCAGTAGCGGCGAGCGAACGGGGAGGAGCCCAGAGCCTGAATCAGTGTGTGTGTTAGTGGAAGCGTCTGGAAAGGCGCGCGATACAGGGTGACAGCCCCGTACACAAAAATGCACATGCTGTGAGCTCGATGAGTAGGGCGGGACACGTGGTATCCTGTCTGAATATGGGGGGACCATCCTCCAAGGCTAAATACTCCTGACTGACCGATAGTGAACCAGTACCGTGAGGGAAAGGCGAAAAGAACCCCGGCGAGGGGAGTGAAAAAGAACCTGAAACCGTGTACGTACAAGCAGTGGGAGCATGCTTAGGCGTGTGACTGCGTACCTTTTGTATAATGGGTCAGCGACTTATATTCTGTAGCAAGGTTAACCGAATAGGGGAGCCGAAGGGAAACCGAGTCTTAACTGGGCGTTAAGTTGCAGGGTATAGACCCGAAACCCGGTGATCTAGCCATGGGCAGGTTGAAGGTTGGGTAACACTAACTGGAGGACCGAACCGACTAATGTTGAAAAATTAGCGGATGACTTGTGGCTGGGGGTGAAAGGCCAATCAAACCGGGAGATAGCTGGTTCTCCCCGAAAGCTATTTAGGTAGCGCCTCGTGAACTCATCTCCGGGGGTAGAGCACTGTTTCGGCAAGGGGGTCATCCCGACTTACCAACCCGATGCAAACTGCGAATACCGGAGAATGTTATCACGGGAGACACACGGCGGGTGCTAACGTCCGTCGNGAAGAGGGAAACAACCCAGACCGCCAGCTAAGGTCCCAAAGTCATGGTTAAGTGGGAAACGATGTGGGAAGGCCCAGACAGCCAGGATGTTGGCTTAGAAGCAGCCATCATTTAAAGAAAGCGTAATAGCTCACTGGTCGAGTCGGCCTGCGCGGAAGATGTAACGGGGCTAAACCATGCACCGAAGCTGCGGCAGCGACGCTTATGCGTTGTTGGGTAGGGGAGCGTTCTGTAAGCCTGTGAAGGTGTACTGTGAGGTATGCTGGAGGTATCAGAAGTGCGAATGCTGACATAAGTAACGATAAAGCGGGTGAAAAGCCCGCTCGCCGGAAGACCAAGGGTTCCTGTCCAACGTTAATCGGGGCAGGGTGAGTCGACCCCTAAGGCGAGGCCGAAAGGCGTAGTCGATGGGAAACAGGTTAATATTCCTGTACTTGGTGTTACTGCGAAGGGGGGACGGAGAAGGCTATGTTGGCCGGGCGACGGTTGTCCCGGTTTAAGCGTGTAGGCTGGTTTTCCAGGCAAATCCGGAAAATCAAGGCTGAGGCGTGATGACGAGGCACTACGGTGCTGAAGCAACAAATGCCCTGCTTCCAGGAAAAGCCTCTAAGCATCAGGTAACATCAAATCGTACCCCAAACCGACACAGGTGGTCAGGTAGAGAATACCAAGGCGCTTGAGAGAACTCGGGTGAAGGAACTAGGCAAAATGGTGCCGTAACTTCGGGAGAAGGCACGCTGATATGTAGGTGAAGCGACTTGCTCGTGGAGCTGAAATCAGTCGAAGATACCAGCTGGCTGCAACTGTTTATTAAAAACACAGCACTGTGCAAACACGAAAGTGGACGTATACGGTGTGACGCCTGCCCGGTGCCGGAAGGTTAATTGATGGGGTTAGCGCAAGCGAAGCTCTTGATCGAAGCCCCGGTAAACGGCGGCCGTAACTATAACGGTCCTAAGGTAGCGAAATTCCTTGTCGGGTAAGTTCCGACCTGCACGAATGGCGTAATGATGGCCAGGCTGTCTCCACCCGAGACTCAGTGAAATTGAACTCGCTGTGAAGATGCAGTGTACCCGCGGCAAGACGGAAAGACCCCGTGAACCTTTACTATAGCTTGACACTGAACATTGAGCCTTGATGTGTAGGATAGGTGGGAGGCTTTGAAGTGTGGACGCCAGTCTGCATGGAGCCGACCTTGAAATACCACCCTTTAATGTTTGATGTTCTAACGTTGACCCGTAATCCGGGTTGCGGACAGTGTCTGGTGGGTAGTTTGACTGGGGCGGTCTCCTCCTAAAGAGTAACGGAGGAGCACGAAGGTTGGCTAATCCTGGTCGGACATCAGGAGGTTAGTGCAATGGCATAAGCCAGCTTGACTGCGAGCGTGACGGCGCGAGCAGGTGCGAAAGCAGGTCATAGTGATCCGGTGGTTCTGAATGGAAGGGCCATCGCTCAACGGATAAAAGGTACTCCGGGGATAACAGGCTGATACCGCCCAAGAGTTCATATCGACGGCGGTGTTTGGCACCTCGATGTCGGCTCATCACATCCTGGGGCTGAAGTAGGTCCCAAGGGTATGGCTGTTCGCCATTTAAAGTGGTACGCGAGCTGGGTTTAGAACGTCGTGAGACAGTTCGGTCCCTATCTGCCGTGGGCGCTGGAGAACTGAGGGGGGCTGCTCCTAGTACGAGAGGACCGGAGTGGACGCATCACTGGTGTTCGGGTTGTCATGCCAATGGCACTGCCCGGTAGCTAAATGCGGAAGAGATAAGTGCTGAAAGCATCTAAGCACGAAACTTGCCCCGAGATGAGTTCTCCCTGACTCCTTGAGGGTCCTGAAGGAACGTTGAAGACGACGACGTTGATAGGCCGGGTGTGTAAGCGCAGCGATGCGTTGAGCTAACCGGTACTAATGAACCGTGAGGCTTAACCTT"}, "NCBI_taxonomy": {"NCBI_taxonomy_name": "Escherichia coli CFT073", "NCBI_taxonomy_id": "199310", "NCBI_taxonomy_cvterm_id": "36763"}, "protein_sequence": {"accession": "", "sequence": ""}}}}, "ARO_category": {"36380": {"category_aro_name": "determinant of lincosamide resistance", "category_aro_cvterm_id": "36380", "category_aro_accession": "3000241", "category_aro_description": "Enzymes, other proteins or other gene products shown clinically to confer resistance to lincosamide antibiotics."}, "35950": {"category_aro_name": "antibiotic resistant gene variant or mutant", "category_aro_cvterm_id": "35950", "category_aro_accession": "0000031", "category_aro_description": "Resistance to antibiotics is often conferred by single nucleotide polymorphisms (SNPs) and other mutations in target genes."}}, "ARO_name": "Escherichia coli 23S rRNA with mutation conferring resistance to clindamycin", "model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "ARO_id": "41296", "model_name": "Escherichia coli 23S rRNA with mutation conferring resistance to clindamycin", "model_type_id": "40295"}, "2822": {"model_id": "2822", "ARO_accession": "3004176", "model_param": {"blastn_bit_score": {"param_value": "5000", "param_type_id": "41093", "param_type": "BLASTN bit-score", "param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}, "snp": {"param_type": "single resistance variant", "param_value": {"7942": "G2087A", "7943": "A2578U"}, "clinical": {"7942": "G2087A", "7943": "A2578U"}, "param_type_id": "36301", "param_description": "A nucleotide or amino acid substitution that confers elevated resistance to antibiotic(s) relative to wild type. The most common type encoded in the CARD is an amino acid substitution gleaned from the literature with format [wild-type][position][mutation], e.g. R184Q. When present in the associated gene or protein, a single resistance variant confers resistance to an antibiotic drug or drug class. Single resistance variants are used by the protein variant and rRNA mutation models to detect antibiotic resistance from submitted sequences."}}, "ARO_description": "Point mutation in the 23S rRNA of Mycoplasma hominis shown to confer resistance to macrolide type antibiotics", "model_sequences": {"sequence": {"4166": {"dna_sequence": {"fmax": "336166", "fmin": "333282", "accession": "CP011538.1", "strand": "-", "sequence": "ACCAATAAATTTATTAGTACTGGTCAGCTGAATGCATTGCTGCACTTACACCTCCAGCCTATCAACCTCATAGTCTATAAGGAATTTTAAAGGGAATACTAATCTTTGAGGGGGCTTCCCGCTTAGATGCTTTCAGCGGTTATCCCTGCCGCACTTGGCTACCCAGCTATGCTTCTGGCGAAACAACTGGCACACCATCGGTGCGTCCACTCCGGTCCTCTCGTACTAAGAGTAGCTCTCATCAATATTCCAACGCCCACATCAGATAGGAACCAAACTGTCTCACGACGTTTTGAACCCAGCTCGCGTACCGCTTTAATGGGCGAACAGCCCAACCCTTGGAACCGACTCCAGCTCCAGGATGCGATGAGCCGACATCGAGGTGCCAAACCTTCCCGTCGATGTGATCTCTTGGGAAAGATAAGCCTGTTATCCCCGGGGTAGCTTTTATCCGTTGAGCGACGGCCTTTCCACGAAGAACCGCCGGATCACTAAGTCCTGCTTTCGCACCTGCTCGACTTGTAGGTCTCACAGTCAATCACACTTCTACCTTTATGCTCTTAGATACGGTTTCTGACCGTATTGAGTGTAACTTTGAACGCCTCCGTTACCCTTTAGGAGGCGACCGCCCCAGTCAAACTACCCACCACACACTGTCCTCTTCCCGGATAACGGGAACAAGTTAGAAGTTCAATGTAACAAGGGTGGTATTTCAACGGCGACTACTCTTAAACTAGCGTTCAAGCATCAACGTCTCCCACCTATCCTACACATGTTAAATCAAACTCCAATATGAAGTTATAGTAAAGCTCCACGGGGTCTTTTCGTCTAGATGCGGGTCTCCGGCGTCTTCGCCGGAACCATAATTTCACCGAGTCTATTGTCGAGACAGTTAAGAGATAATTACTCCTTTCGTGCAGGTCAGTATTTAGCCGACAAGGAATTTCGCTACCTTAGGACCGTTATAGTTACGGCCGCCGTTCACCCGGGCTTCACATTAATGCTTCGCTAATGCTAACACCTCTGCTTAACCTTCGGGCACTGGGCAGGAGTCACCCCATATACATCATCTTACGACTTAGCATAGAGCTGTGTTTTTGATAAACAGTTCCCCCTTACTATTCACTGCGGCCCACATTTTATTGTGGGCATCCCTTCTTGCGAACTTACGGGATGAATTTGCAGAGTTCCTTGACAATAGTTTTCTCGCTCGCCTTAGAATACTCATCTTGGGGACGTGTGTCCGTTCTCGGTACGGGTTTCCTAACACTTAATGTTAGAAGCTTTTTTAAGAGGCATGAAATCATCTAATTCGCTACTCAGTTGCCCTTTTGCTATGCGTCGTAACTCCCAGTTAAGTTATGCGGATTTGCCTACATAACCCAGTTGTTACTTACCCCACAATCCAATTAGTGGTAAAATTATCCTTCCCCGTCACTCCATCACATGTTAAGAAAGTACAGGAATATTAACCTGTTGTCCATCGGCTACGCCTTTCGGCCTCGTCTTAGGACCCGACTAACCCTGGGTGGACGAACCTTGCCCAGGAAACCTTCCCCAATAGGCGTCAGAGATTCTCACTCTGAATCGTTACTCATACCGGCATTCTCACTTGTAAGCGCTCCACCAGTCCTCACGGTCTGGCTTCTAAGCCCTTACAACGCTCTCCTAACGCATTTCTGCCCGTAGCTTCGGTATTGTGTTTTAGTCCCGTTGAATTATCGGCACAAAGTCTCTCGACTAGTGAGCTATTACGCACTCTTTAAACGGTGGCTGCTTCTAAGCCAACATCCTAGCTGTTTAAGAAACTTCACAACCTTTCTCACTTAACACAATTTTGGGACCTTAGCTGACGATCTGGGTTGTTCCCCTCGCGTGCATCGACGTTATCACCGATGTACCGACTGCATAGTAATACATGATAGTATTCGGAGTTTGATTATAGTCAGTACGGCTAGGCGCCGCCATTCCATATTCAGTGCTCTACCCCCATCATTTAACACTACACGCTAGCCCTAAAGCTATTTCGGAGAGAACCAGCTATCTCCAAGTTCGATTGGAATTTCACCCCTACCCACAAGTCATCCGGGCACTTTTTAGCGTACTGCGGTTCGGTCCTCCACTTAGTGTTACCTAAGTTTCAACCTGCTCATGGGTAGATCACCTGGTTTCGGGTCTATATCAACATACTAAATCGCCCTATTCAGACTCGATTTCTCTACGGCTTCGCTTTATTCTACTTAACCTCGCATGTTGACATAACTCGCCGGTCCATTCTGCAAGATGTACGCCATCACCCATTAACGGGCTCTGACTAACTGTAAGTAATTGGTTTCAGAATCTATTTCACTCCCCTCCCGGGGTTCTTTTCACCTTTCCCTCACGGTACTAGTTCACTATCGGTGTCTGGTTAGTATTTAGCCTTACCGGGTGGTCCCGGCAAATTCAGACAGGGTTTCACGTGCCCCGCCCTACTCAGGATACTGCTAGGAGATTTATACATTTCGCTTACGGGAATTTCACCCTCTATGTTTAAGCGTTCCAACTTATTCTGCTATGTACAAATTCATTAATCCATGTCGCAGTCCTACAACCCCAACAACATGTGTTGGTTTGGGCTCTTCCCCGTTCGCTCGCCACTACTTAGGGAATCATTCTTTATTTTCTCTTCCTGCTGCTACTGAGATGTTTCAATTCACAGCGTGTCTCTTCATTCGACTATGAATTCATCGATATGATAATTGAGGATTAGCTCAATTAGGTTTCCCCATTCGGAAATCCCCGGATAACAGCTTATTTCCAGCTAACCGAGGCTTATCGCAGGTAATCACGTCCTTCATCGACTTCCAGACCCAAGGCATCCACCAAAAACTCTTGCTTGTTT"}, "NCBI_taxonomy": {"NCBI_taxonomy_name": "Mycoplasma hominis", "NCBI_taxonomy_id": "2098", "NCBI_taxonomy_cvterm_id": "40311"}, "protein_sequence": {"accession": "", "sequence": ""}}}}, "ARO_category": {"36454": {"category_aro_name": "determinant of macrolide resistance", "category_aro_cvterm_id": "36454", "category_aro_accession": "3000315", "category_aro_description": "Enzymes, other proteins or other gene products shown clinically to confer resistance to macrolide antibiotics."}, "35950": {"category_aro_name": "antibiotic resistant gene variant or mutant", "category_aro_cvterm_id": "35950", "category_aro_accession": "0000031", "category_aro_description": "Resistance to antibiotics is often conferred by single nucleotide polymorphisms (SNPs) and other mutations in target genes."}}, "ARO_name": "Mycoplasma hominis 23S rRNA with mutation conferring resistance to macrolide antibiotics", "model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "ARO_id": "41327", "model_name": "Mycoplasma hominis 23S rRNA with mutation conferring resistance to macrolide antibiotic", "model_type_id": "40295"}, "2823": {"model_id": "2823", "ARO_accession": "3004177", "model_param": {"blastn_bit_score": {"param_value": "5000", "param_type_id": "41093", "param_type": "BLASTN bit-score", "param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}, "snp": {"param_type": "single resistance variant", "param_value": {"7944": "A2021A"}, "clinical": {"7944": "A2021A"}, "param_type_id": "36301", "param_description": "A nucleotide or amino acid substitution that confers elevated resistance to antibiotic(s) relative to wild type. The most common type encoded in the CARD is an amino acid substitution gleaned from the literature with format [wild-type][position][mutation], e.g. R184Q. When present in the associated gene or protein, a single resistance variant confers resistance to an antibiotic drug or drug class. Single resistance variants are used by the protein variant and rRNA mutation models to detect antibiotic resistance from submitted sequences."}}, "ARO_description": "Point mutation in the 23S rRNA of Mycoplasma fermentans shown to confer resistance to macrolide type antibiotics", "model_sequences": {"sequence": {"4167": {"dna_sequence": {"fmax": "904756", "fmin": "901870", "accession": "NC_021002.1", "strand": "-", "sequence": "CTTTTAAACCGATCGATTTATTAGTATTGGTCAGCTCAACGTATTACTACGCTTACACCTCCAACCTATCAACCTCATAGTCTATAAGGAATCTCAAGGGAATACTTATCTCTGAGGAGGCTTCCCACTTAGATGCTTTCAGCGGTTATCCTTTCCGTACTTAGCTACCCAGCTATGCTCCTGGCGGAACAACTGGAACACCAGTGGTACGTCCACTCCGGTCCTCTCGTACTAAGAGCGGCTCTCATCAATATTCCAACGCCCACATCAGATAGGGACCAAACTGTCTCACGACGTTTTGAACCCAGCTCGCGTACCGCTTTAATGGGCGAACAGCCCAACCCTTGGAACCGACTCCAGCTCCAGGATGCGATGAGCCGACATCGAGGTGCCAAACCTTGCCGTCGATGTGATCTCTTGGGCAAGATAAGCCTGTTATCCCCAGGGTAACTTTTATCCGTTGAGCGACTGCCATTCCACAATGTACAGCCGGATCACTAAGTCCTGCTTTCGCACCTGCTCGACTTGTAAGTCTCACAGTCAAGCACACTTCTACCTTTGCGCTCTACATACGGTTTCTGACCGTACTGAGTGTACCTTTGAACGCCTCCGTTACTCTTTAGGAGGCGACCGCCCCAGTCAAACTACCCGTCACGCACTGTCCACCCACCTGATGATGATGGCATGTTAGAAACTCAATATACCAAGGGTGGTATTTCAAGGATGACTCCACGAGAACTAGCGTCCTCGCTTCAAAGTCTCCCACCTATCCTACACATGTTAGACCAAGTTCCAATACGAAACTGTAGTAAAGCTCCATGGGGTCTTTTCGTCTTGATGCGGGTACCCAGCGTTTTCACTGGGACCATAATTTCACCGAGTCTAGTGTTGAGACAGTTGAGAGATCATTGCGCCTTTCGTGCAGGTCAGTATTTAGCCGACAAGGAATTTCGCTACCTTAGGACCGTTATAGTTACGGCCGCCGTTCACCCGGGCTTCATTTCAACGCTTCGCAAATGCTAACGCATCCACTTAACCTTCGGGCACTGGGCAGGCTTCACCCCCTATACATCACCTTGCGGTTTAGCAGAGAGCTGTGTTTTTGATAAACAGTTGCCCCTCATAATTTACTGTGGCCTAAGTGTTACCTTAGGCGCCCCTTCTTGCGAACTTACGGGGTCATTTTGCAGAGTTCCTTAACACTAGTTTTCTCGCTCGCCTTAGAATACTCATCTCGGGGACGTGTGTCCGTTCTCGGTACAGGTTTCCATAATATTAAGTTTAGAAGCTTTTCTTGGAAGTGTGGAATCATCTAATTCGGTTTGACCCTATGCATCACGCCTCCCGGTTATAGACTGCGGATTTGCCTACAGTCACCAGTGAACGCTTACCCCACAATCCAGTAAGTGGTAAGATTATCCTCCTCCGTCACTCCATCACTATTATAGAAAGTACAGGAATATTAACCTGTTGTCCATCGGCTACGCTTTTCAGCCTCGTCTTAGGTCCTGACTAACCCTGGGTGGACGAACCTTGCCCAGGAAACCTTCCCCAATAGGCGTCGTAGATTCTCACTACGAATCGTTACTCATACCGGCATTCTCACTTCCTAGCGCTCCACCAGTCCTCACGGTCTGACTTCATTGCCCTAGGAACGCTCCTCTAACGTAAATATAATTACCCGCGGCTTCGGTATCGTGTTTTAGTCCCGTTAAATTGTTGGCGCAAGGTCTCTTGACTAGTGAGCTATTACGCACTCTTTAAAAGGTGGCTGCTTCTGAGCCAACTTCCTAGTTGTTTATGAAACCTCACAACCTTTCTCACTTAACACGATTTTGGGACCTTAGCCGGCGATCTGGGTTGTTGCCCTCGCGAGCCGGGACGTTAGCACCCCGGTTCCGACTGCATAGCAATACATAGCGGTATTCGGAGTTTGATTATAGTCAGTACCCCTAGGCGAGGCCATTCCATATTCAGTGCTCTACCACCACTACTTAACACTACACGCTAGCCCTAAAGCTATTTCGAGGAGAACCAGCTATCTCCAAGTTCGATTGGAATTTCTCCACTATTCACAAGTCATCCGGGCACTTTTCAGCGTACTACGGTTCGGCCCTCCGCTTGGGGTTAGCCAAGTTTCAGCCTGCTCATGAATAGATCACATGGTTTCGGGTATATGACAACATACTAAGACGCCCTATTAAGACTCGATTTCTCTACGGCTCCGCTTTTTTCTGCTTAACCTCGCATGCTGTCATAACTCGCCGGTCCATACTGCAAGATGTACGCCATCACCCATAAACGGGCTCTGACTAATTGTAAGTAAGTGGTTTCAGAATCTATTTCACTCCCCTCTCGGGGTTCTTTTCACCTTTCCCTCACGGTACTAGTTCACTATCGGTGTCTGATTAGTATTTAGCCTTACCGGGTGGTCCCGGCAGATTCAGACAGGGTTCCACGTGCCCCGCCCTACTCAGGATACGATCAGAAGACTTAACAATTTCACATACGGGGGTATCACCCTCTATGCCGCTTCTTCCCAAAAGCTTCTGCTATCATTAAGTTTTGTAACTTCATGTAGATCGTCCTACAACCCCCATTGCTGGGTTTGGGCTCTTTCTCGTTCGCTCGCCGCTACTAAAGAAATCATTGTTTATTTTCTCTTCCTCTTGCTACTAAGATGTTTCAGTTCACAAGGTGTCTCACTCAAGTTCCTATGAATTCAGAACAAGGCAACTAGGCATTACCCTAGTTAGGTTTCCCCATTCGGAAATCCCCGTTTCATAGCATATTTCCGGCTCCACGAGGCTTATCGCAGGTAATCACGTCCTTCATCGACTTTCAGACCCAAGGCATCCACCACAAACTCTTCCTTATTTAAAAGTA"}, "NCBI_taxonomy": {"NCBI_taxonomy_name": "Mycoplasma fermentans PG18", "NCBI_taxonomy_id": "496833", "NCBI_taxonomy_cvterm_id": "41329"}, "protein_sequence": {"accession": "", "sequence": ""}}}}, "ARO_category": {"36454": {"category_aro_name": "determinant of macrolide resistance", "category_aro_cvterm_id": "36454", "category_aro_accession": "3000315", "category_aro_description": "Enzymes, other proteins or other gene products shown clinically to confer resistance to macrolide antibiotics."}, "35950": {"category_aro_name": "antibiotic resistant gene variant or mutant", "category_aro_cvterm_id": "35950", "category_aro_accession": "0000031", "category_aro_description": "Resistance to antibiotics is often conferred by single nucleotide polymorphisms (SNPs) and other mutations in target genes."}}, "ARO_name": "Mycoplasma fermentans 23S rRNA with mutation conferring resistance to macrolide antibiotics", "model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "ARO_id": "41328", "model_name": "Mycoplasma fermentans 23S rRNA with mutation conferring resistance to macrolide antibiotics", "model_type_id": "40295"}, "2815": {"model_id": "2815", "ARO_accession": "3004168", "model_param": {"blastn_bit_score": {"param_value": "5700", "param_type_id": "41093", "param_type": "BLASTN bit-score", "param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}, "snp": {"param_type": "single resistance variant", "param_value": {"8171": "A2265U"}, "clinical": {"8171": "A2265U"}, "param_type_id": "36301", "param_description": "A nucleotide or amino acid substitution that confers elevated resistance to antibiotic(s) relative to wild type. The most common type encoded in the CARD is an amino acid substitution gleaned from the literature with format [wild-type][position][mutation], e.g. R184Q. When present in the associated gene or protein, a single resistance variant confers resistance to an antibiotic drug or drug class. Single resistance variants are used by the protein variant and rRNA mutation models to detect antibiotic resistance from submitted sequences."}}, "ARO_description": "Point mutation in the 23S rRNA of Mycobacterium kansasii shown to confer resistance to clarithromycin, a macrolide type antibiotic.", "model_sequences": {"sequence": {"4178": {"dna_sequence": {"fmax": "5021505", "fmin": "5018387", "accession": "NZ_CP019883.1", "strand": "+", "sequence": "TTGTAAGTGTCTAAGGGCGCATGGTGGATGCCTTGGCATCGAGAGCCGATGAAGGACGTGGGAGGCTGCGATAAGCCTCGGGGAGCTGTCAACCGAGCGTGGATCCGAGGATTTCCGAATGGGGAAACCCAGCACGAGTGATGTCGTGTTACCCGCATCTGAATATATAGGGTGCGGGAGGGAACGCGGGGAAGTGAAACATCTCAGTACCCGTAGGAGAAGAAAACAAAAGTGATTCCGTAAGTAGTGGCGAGCGAACGCGGAACATGGCTAAACCGCACGCATGGGTAACCGGGTAGGGGTTGTGTGTGCGGGGTTGTGGGATCGATACGTCTCAGCTCTACCCGGCTGAGGGGCAGTCAGAAAGTGTCGTGGTTAACGGAAGTGGCCTGGGATGGTCTGCCGTAGACGGTGAGAGCCCGGTACGTGAAAACCCGGCACCTGCCTTGTATCAATTCCCGAGTAGCAGCGGGCCCGTGGAATCTGCTGTGAATCTGCCGGGACCACCCGGTAAGCCTAAATACTCCTCGATGACCGATAGCGGAATAGTACCGTGAGGGAATGGTGAAAAGTACCCCGGGAGGGGAGTGAAAGAGTACCTGAAACCGTGTGCCTACAATCCGTCAGAGCCCTTTCGTGGGGTGATGGCGTGCCTTTTGAAGAATGAGCCTGCGAGTCAGGGACATGTCGCGAGGTTAACCCGTGCGGGGTAGCCGCAGCGAAAGCGAGTCTGAATAGGGCGTATCGCGCGCGAGCGTGTGTAGTGGCGTGTTCTGGACCCGAAGCGGAGTGATCTACCCATGGCCAGGGTGAAGCGCGGGTAAGACCGCGTGGAGGCCCGAACCCACTTAGGTTGAAGACTGAGGGGATGAGCTGTGGGTAGGGGTGAAAGGCCAATCAAACTCCGTGATAGCTGGTTCTCCCCGAAATGCATTTAGGTGCAGCGTTGCGTGTTTCACCACGGAGGTAGAGCTACTGGATGGCCGATGGGCCCCACTAGGTTACTGACGTCAGCCAAACTCCGAATGCCGTGGTGTATAGCGTGGCAGTGAGACGGCGGGGGATAAGCTCCGTACGTCGAAAGGGAAACAGCCCAGATCGCCGGCTAAGGCCCCAAAGCGTGTGCTAAGTGGGAAAGGATGTGCAGTCGCAGAGACAACCAGGAGGTTGGCTTAGAAGCAGCCACCCTTGAAAGAGTGCGTAATAGCTCACTGGTCAAGTGATTGTGCGCCGATAATGTAGCGGGGCTCAAGCACACCGCCGAAGCCGCGACAACCGCAAGGTTGGGTAGGGGAGCGTCCCTCATTCAGCGAAGCTGCCGGGTGACCGGTGGTGGAGGATGGGGGAGTGAGAATGCAGGCATGAGTAGCGATAAGGCAAGTGAGAACCTTGCCCGCCGAAAGACCAAGGGTTCCTGGGCCAGGCCAGTCCGCCCAGGGTGAGTCGGGACCTAAGGCGAGGCCGACAGGCGTAGTCGATGGACAACGGGTTGATATTCCCGTACCCGTGTGTGGGCGCCCGTGATGAATCAGCGGTACTAACCACCCAAAACCGGATCGATCACTCCCCTTCGGGGGCGTGGAGGTCTGGGGCTGCGTGGAGCCTTCGCTGGTAGTAGTCAAGCGATGGGGTGACGCAGGAAGGCAGCCGTACCAGTCAGTGGTAATACTGGGGCAAGCCAGTAGGGAGAGCGATAGGCAAATCCGTCGCTCACAAATCCTGAGAGGTGACGCATAGCCGATTGAGGCGAATTCGGTGATCCTCTGCTGCCAAGAAAAGCCTCTAGCGAGCACACACACGGCCCGTACCCCAAACCGACACAGGTGGTCAGGTAGAGAATACCAAGGCGTACGAGATAACTATGGTTAAGGAACTCGGCAAAATGCCCCCGTAACTTCGGGAGAAGGGGGACCGGAATACCGTGAACACCCTTGCGGTGGGAGCGGGATTCGGTCGCAGAAACCAGTGAGAAGCGACTGTTTACTAAAAACACAGGTCCGTGCGAAGTCGCAAGACGATGTATACGGACTGACGCCTGCCCGGTGCTGGAAGGTTAAGAGGACCCGTTAACCCGCAAGGGTGAAGCGGAGAATTTAAGCCCCAGTAAACGGCGGTGGTAACTATAACCATCCTAAGGTAGCGAAATTCCTTGTCGGGTAAGTTCCGACCTGCACGAATGGCGTAACGACTTCTCAACTGTCTCAACCATAGACTCGGCGAAATTGCACTACGAGTAAAGATGCTCGTTACGCGCGGCAGGACGAAAAGACCCCGGGACCTTCACTACAACTTGGTATTGGTGTTCGGTACGGTTTGTGTAGGATAGGTGGGAGACTGTGAAACCTCAACGCCAGTTGGGGTGGAGTCGTTGTTGAAATACCACTCTGATCGTATTGGACACCTAACGTCGAACCCTGAATCGGGTTCACGGACAGTGCCTGGCGGGTAGTTTAACTGGGGCGGTTGCCTCCTAAAATGTAACGGAGGCGCCCAAAGGTTCCCTCAACCTGGACGGCAATCAGGTGGCGAGTGTAAGTGCACAAGGGAGCTTGACTGCGAGACCTACAAGTCAAGCAGGGACGAAAGTCGGGACTAGTGATCCGGCACCTCTGAGTGGAAGGGGTGTCGCTCAACGGATAAAAGGTACCCCGGGGATAACAGGCTGATCTTCCCCAAGAGTCCATATCGACGGGATGGTTTGGCACCTCGATGTCGGCTCGTCGCATCCTGGGGCTGGAGCAGGTCCCAAGGGTTGGGCTGTTCGCCCATTAAAGCGGCACGCGAGCTGGGTTTAGAACGTCGTGAGACAGTTCGGTCTCTATCCGCCGCGCGCGTCAGAAGCTTGAGGAAACCTGTCCCTAGTACGAGAGGACCGGGACGGACGAACCTCTAGTGCACCAGTTGTCCCACCAGGGGCACCGCTGGATAGCTACGTTCGGACAGGATAACCGCTGAAAGCATCTAAGCGGGAAACCTTCTCCAAGATCAGGCTTCTCACCCACTTGGTGGGATAAGGCCCCCCGCAGAACACGGGTTCGATAGGCCAGACCTGGAAGCTCAGTAATGAGTGAAGGGAACTGGCACTAACCGGCCGAAAACTTACCAACACAAATAATCG"}, "NCBI_taxonomy": {"NCBI_taxonomy_name": "Mycobacterium kansasii", "NCBI_taxonomy_id": "1768", "NCBI_taxonomy_cvterm_id": "41347"}, "protein_sequence": {"accession": "", "sequence": ""}}}}, "ARO_category": {"35950": {"category_aro_name": "antibiotic resistant gene variant or mutant", "category_aro_cvterm_id": "35950", "category_aro_accession": "0000031", "category_aro_description": "Resistance to antibiotics is often conferred by single nucleotide polymorphisms (SNPs) and other mutations in target genes."}, "36454": {"category_aro_name": "determinant of macrolide resistance", "category_aro_cvterm_id": "36454", "category_aro_accession": "3000315", "category_aro_description": "Enzymes, other proteins or other gene products shown clinically to confer resistance to macrolide antibiotics."}}, "ARO_name": "Mycobacterium kansasii 23S rRNA with mutation conferring resistance to clarithromycin", "model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "ARO_id": "41318", "model_name": "Mycobacterium kansasii 23S rRNA with mutation conferring resistance to clarithromycin", "model_type_id": "40295"}, "2814": {"model_id": "2814", "ARO_accession": "3004167", "model_param": {"blastn_bit_score": {"param_value": "5700", "param_type_id": "41093", "param_type": "BLASTN bit-score", "param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}, "snp": {"param_type": "single resistance variant", "param_value": {"7923": "A2268C"}, "clinical": {"7923": "A2268C"}, "param_type_id": "36301", "param_description": "A nucleotide or amino acid substitution that confers elevated resistance to antibiotic(s) relative to wild type. The most common type encoded in the CARD is an amino acid substitution gleaned from the literature with format [wild-type][position][mutation], e.g. R184Q. When present in the associated gene or protein, a single resistance variant confers resistance to an antibiotic drug or drug class. Single resistance variants are used by the protein variant and rRNA mutation models to detect antibiotic resistance from submitted sequences."}}, "ARO_description": "Point mutation in the 23S rRNA of Mycobaccterium avium shown to confer resistance to azithromycin, a macrolide type antibiotic.", "model_sequences": {"sequence": {"4159": {"dna_sequence": {"fmax": "3103", "fmin": "0", "accession": "NR_076151.1", "strand": "+", "sequence": "TAAGTGTTTAAGGGCGCATGGTGGATGCCTTGGCATCGAGAGCCGATGAAGGACGTGGGAGGCTGCGATATGCCTCGGGGAGCTGTCAACCGAGCATTGATCCGAGGATTTCCGAATGGGGAAACCCAGCACGAGTGATGTCGTGTTACCCGCATCTGAATATATAGGGTGCGGGAGGGAACGCGGGGAAGTGAAACATCTCAGTACCCGTAGGAGAAGAAAACAATTGTGATTCCGTAAGTAGTGGCGAGCGAACGCGGAACAGGCTAAACCGCACGCATGTGATACCGGGTAGGGGTTGTGTGTGCGGGGTTGTGGGAGGATACATCTCAGCTCTACCTGGCTGAGGGGTAGTCAGAAAGTGTCGTGGTTAGCGGAAGTGGCCTGGGATGGTCTGCCGTAGACGGTGAGAGCCCGGTACGCGAAAACCCGTCACCTACCTTGTATCAATTCCCGAGTAGCAGCGGGCCCGTGGAATCTGCTGTGAATCTGCCGGGACCACCCGGTAAGCCTAAATACTTCTCGATGACCGATAGCGGATTAGTACCGTGAGGGAATGGTGAAAAGTACCCCGGGAGGGGAGTGAAATAGTACCTGAAACCGTGTGCCTACAATCCGTCAGAGCCTCCTTGTGGGGTGATGGCGTGCCTTTTGAAGAATGAGCCTGCGAGTCAGGGACACGTCGCGAGGTTAACCCGTGCGGGGTAGCCGCAGCGAAAGCGAGTCTGAATAGGGCGCATCCCCTTTGGGGTGTAGTGGCGTGTTCTGGACCCGAAGCGGAGTGATCTACCCATGGCCAGGGTGAAGCGCGGGTAAGACCGCGTGGAGGCCCGAACCCACTTAGGTTGAAGACTGAGGGGATGAGCTGTGGGTAGGGGTGAAAGGCCAATCAAACTCCGTGATAGCTGGTTCTCCCCGAAATGCATTTAGGTGCAGCGTTGCGTGGTTCACCACGGAGGTAGAGCTACTGGATGGCCGATGGGCCCTACTAGGTTACTGACGTCAGCCAAACTCCGAATGCCGTGGTGTAAAGCGTGGCAGTGAGACGGCGGGGGATAAGCTCCGTACGTCGAAAGGGAAACAGCCCAGATCGCCGGCTAAGGCCCCTAAGCGTGTGCTAAGTGGAAAAGGATGTGTAGTCGCAGAGACAACCAGGAGGTTGGCTTAGAAGCAGCCACCCTTGAAAGAGTGCGTAATAGCTCACTGGTCAAGTGATTATGCGCCGATAATGTAGCGGGGCTCAAGCACACCGCCGAAGCCGCGGCACATTCACGTTTACGTGGATGTGGGTAGGGGGAGCGTCCCTCATTCAGCGAAGCCTCCGGGTGACCGGTGGTGGAGGGTGGGGGAGTGAGAATGCAGGCATGAGTAGCGATAAGGCAAGTGAGAACCTTGCCCGCCGTAAGACCAAGGGTTCCTGGGCCAGGCCAGTCCGCCCAGGGTGAGTCGGGACCTAAGGCGAGGCCGACAGGCGTAGTCGATGGACAACGGGTTGATATTCCCGTACCCGTGTATGGGCGTCCCTGATGAATCAGCGGTACTAACCACCCAAAACCGGATCGACCATTCCCCTTCGGGGGCATGGAGTTTCGGGGCTGCGTGGGACCTTCGCTGGTAGTAGTCAAGCAATGGGGTGACGCAGGAAGGTAGCCGTACCAGTCAGTGGTAATACTGGGGCAAGCCTGTAGGGAGAGCGATAGGCAAATCCGTCGCTCATTAATCCTGAGAGGTGATGCATAGCCGATTGAGGTGAATTCGGTGATCCTCTGCTGCCAAGAAAAGCCTCTAGCGAGCACATACACGGCCCGTACCCCAAACCAACACAGGTGGTCAGGTAGAGAATACCAAGGCGTACGAGATAACTATGGTTAAGGAACTCGGCAAAATGCCCCCGTAACTTCGGGAGAAGGGGGGCCGGAATACCGTGAACACCCTTGCGGTGGGAGCGGGATCCGGCCGCAGAAACCAGTGGGTAGCGACTGTTTACTAAAAACACAGGTCCGTGCGAAGTCGCAAGACGATGTATACGGACTGACGCCTGCCCGGTGCTGGAAGGTTAAGAGGACCCGTTAACCGTAAGGTGAAGCGGAGAATTTAAGCCCCAGTAAACGGCGGTGGTAACTATAACCATCCTAAGGTAGCGAAATTCCTTGTCGGGTAAGTTCCGACCTGCACGAATGGCGTAACGACTTCCCAACTGTCTCAACCATAGACTCGGCGAAATTGCACTACGAGTAAAGATGCTCGTTACGCGCGGCAGGACGAAAAGACCCCGGGACCTTCACTACAACTTGGTATTGGTGTTCGGTACGGTTTGTGTAGGATAGGTGGGAGACTGTGAAATACAGACGCCAGTTTGTATGGAGTCGTTGTTGAAATACCACTCTGATCGTATTGGACACCTAACGTCGAACCCTTATCGGGTTCACGGACAGTGCCTGGCGGGTAGTTTAACTGGGGCGGTTGCCTCCTAAAATGTAACGGAGGCGCCCAAAGGTTCCCTCAACCTGGACGGCAATCAGGTGACGAGTGTAAGTGCACAAGGGAGCTTGACTGCGAGACTTACAAGTCAAGCAGGGACGAAAGTCGGGACTAGTGATCCGGCACCCCCGAGTGGAAGGGGTGTCGCTCAACGGATAAAAGGTACCCCGGGGATAACAGGCTGATCTTCCCCAAGAGTCCATATCGACGGGATGGTTTGGCACCTCGATGTCGGCTCGTCGCATCCTGGGGCTGGAGCAGGTCCCAAGGGTTGGGCTGTTCGCCCATTAAAGCGGCACGCGAGCTGGGTTTAGAACGTCGTGAGACAGTTCGGTCTCTATCCGCCGCGCGCGTCAGAAACTTGAGGAAACCTGTCCCTAGTACGAGAGGACCGGGACGGACGAACCTCTGGTATACCAGTTGTTCCACCAGGAGCACGGCTGGATAGCCACGTTCGGACAGGATAACCGCTGAAAGCATCTAAGCGGGAAACCTTCTCCAAGATCAGGTTTCTCACCCTTTTAGAGGGATAAGGCCCCCCGCAGACCACGGGTTCGATAGGCTAGACCTGGAAGCTCAGCAATGAGTGCAGGGAACTGGCACTAACCGGCCGAAAACTTAC"}, "NCBI_taxonomy": {"NCBI_taxonomy_name": "Mycobacterium intracellulare", "NCBI_taxonomy_id": "1767", "NCBI_taxonomy_cvterm_id": "41317"}, "protein_sequence": {"accession": "", "sequence": ""}}}}, "ARO_category": {"35950": {"category_aro_name": "antibiotic resistant gene variant or mutant", "category_aro_cvterm_id": "35950", "category_aro_accession": "0000031", "category_aro_description": "Resistance to antibiotics is often conferred by single nucleotide polymorphisms (SNPs) and other mutations in target genes."}, "36454": {"category_aro_name": "determinant of macrolide resistance", "category_aro_cvterm_id": "36454", "category_aro_accession": "3000315", "category_aro_description": "Enzymes, other proteins or other gene products shown clinically to confer resistance to macrolide antibiotics."}}, "ARO_name": "Mycobacterium intracellulare 23S rRNA with mutation conferring resistance to azithromycin", "model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "ARO_id": "41316", "model_name": "Mycobacterium intracellulare 23S rRNA with mutation conferring resistance to azithromycin", "model_type_id": "40295"}, "2817": {"model_id": "2817", "ARO_accession": "3004170", "model_param": {"blastn_bit_score": {"param_value": "5000", "param_type_id": "41093", "param_type": "BLASTN bit-score", "param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}, "snp": {"param_type": "single resistance variant", "param_value": {"8178": "C2641G", "8175": "A2125G", "8176": "A2126G", "8177": "C2641A"}, "clinical": {"8178": "C2641G", "8175": "A2125G", "8176": "A2126G", "8177": "C2641A"}, "param_type_id": "36301", "param_description": "A nucleotide or amino acid substitution that confers elevated resistance to antibiotic(s) relative to wild type. The most common type encoded in the CARD is an amino acid substitution gleaned from the literature with format [wild-type][position][mutation], e.g. R184Q. When present in the associated gene or protein, a single resistance variant confers resistance to an antibiotic drug or drug class. Single resistance variants are used by the protein variant and rRNA mutation models to detect antibiotic resistance from submitted sequences."}}, "ARO_description": "Point mutation in the 23S rRNA of Streptococcus pneumoniae shown to confer resistance to macrolide type antibiotics", "model_sequences": {"sequence": {"4179": {"dna_sequence": {"fmax": "1695312", "fmin": "1692408", "accession": "NZ_CP018138.1", "strand": "+", "sequence": "TTTGGATAAGTCCTCGAGCTATTAGTATTAGTCCGCTACATGTGTCGCCACACTTCCACTTCTAACCTATCTACCTGATCATCTCTCAGGGCTCTTACTGATATATAATCATGGGAAATCTCATCTTGAGGTGGGTTTCACACTTAGATGCTTTCAGCGTTTATCCCTTCCCTACATAGCTACCCAGCGATGCCTTTGGCAAGACAACTGGTACACCAGCGGTAAGTCCACTCTGGTCCTCTCGTACTAGGAGCAGATCCTCTCAAATTTCCTACGCCCGCGACGGATAGGGACCGAACTGTCTCACGACGTTCTGAACCCAGCTCGCGTGCCGCTTTAATGGGCGAACAGCCCAACCCTTGGGACCGACTACAGCCCCAGGATGCGACGAGCCGACATCGAGGTGCCAAACCTCCCCGTCGATGTGAACTCTTGGGGGAGATAAGCCTGTTATCCCCAGGGTAGCTTTTATCCGTTGAGCGATGGCCCTTCCATACGGAACCACCGGATCACTAAGCCCGACTTTCGTCCCTGCTCGAGTTGTAGCTCTCGCAGTCAAGCTCCCTTATACCTTTATACTCTGCGAATGATTTCCAACCATTCTGAGGGAACCTTTGGGCGCCTCCGTTACCTTTTAGGAGGCGACCGCCCCAGTCAAACTGCCCGTCAGACACTGTCTCCGATAGGGATCACCTATCTGGGTTAGAGTGGCCATAACACAAGGGTAGTATCCCAACAGCGTCTCCTTCGAAACTGGCGTCCCGATCTCTTAGACTCCTACCTATCCTGTACATGTGGTACAGACACTCAATATCAAACTGCAGTAAAGCTCCATGGGGTCTTTCCGTCCTGTCGCGGGTAACCTGCATCTTCACAGGTACTAAAATTTCACCGAGTCTCTCGTTGAGACAGTGCCCAAATCATTACGCCTTTCGTGCGGGTCGGAACTTACCCGACAAGGAATTTCGCTACCTTAGGACCGTTATAGTTACGGCCGCCGTTTACTGGGGCTTCAATTCATACCTTCGCTTACGCTAAGCACTCCTCTTAACCTTCCAGCACCGGGCAGGCGTCACCCCCTATACATCATCTTACGATTTAGCAGAGAGCTGTGTTTTTGATAAACAGTTGCTTGGGCCTATTCACTGCGGCTGACCTAAAGTCAGCACCCCTTCTCCCGAAGTTACGGGGTCATTTTGCCGAGTTCCTTAACGAGAGTTCTCTCGCTCACCTGAGGCTACTCGCCTCGACTACCTGTGTCGGTTTGCGGTACGGGTAGAGTATGTTTAAACGCTAGAAGCTTTTCTTGGCAGTGTGACGTCACTAACTTCGCTACTAAACTTCGCTCCCCATCACAGCTCAATGTTATAGAATTAAGCATTTGACTCAATTCACACCTCACTGCTTAGACAGACTCTTCCAATCGTCTGCTTTAGTTAGCCTACTGCGTCCCTCCATCACTACATACTCTAGTACAGGAATATCAACCTGTTGTCCATCGGATACACCTTTCGGTCTCTCCTTAGGTCCCGACTAACCCAGGGCGGACGAGCCTTCCCCTGGAAACCTTAGTCTTACGGTGGACAGGATTCTCACCTGTCTTTCGCTACTCATACCGGCATTCTCACTTCTATGCGTTCCAGCACTCCTCACGGTATACCTTCATCACACATAGAACGCTCTCCTACCATACCTATAAAGGTATCCACAGCTTCGGTAAATTGTTTTAGCCCCGGTACATTTTCGGCGCAGGGTCACTCGACTAGTGAGCTATTACGCACTCTTTGAATGAATAGCTGCTTCTAAGCTAACATCCTAGTTGTCTGTGCAACCCCACATCCTTTTCCACTTAACAATTATTTTGGGACCTTAGCTGGTGGTCTGGGCTGTTTCCCTTTCGACTACGGATCTTAGCACTCGCAGTCTGACTGCCGACCATAATTCATTGGCATTCGGAGTTTATCTGAGATTGGTAATCCGGGATGGACCCCTCACCCAAACAGTGCTCTACCTCCAAGAATCTCTAATGTCGACGCTAGCCCTAAAGCTATTTCGGAGAGAACCAGCTATCTCCAAGTTCGTTTGGAATTTCTCCGCTACCCACAAGTCATCCAAGCACTTTTCAACGTGCCCTGGTTCGGTCCTCCAGTGCGTCTTACCGCACCTTCAACCTGCTCATGGGTAGGTCACATGGTTTCGGGTCTACGTCATGATACTAATTCGCCCTGTTCAGACTCGGTTTCCCTACGGCTCCGTCTCTTCAACTTAACCTCGCATCATAACGTAACTCGCCGGTTCATTCTACAAAAGGCACGCTCTCACCCATTAACGGGCTCGAACTTGTTGTAGGCACACGGTTTCAGGTTCTATTTCACTCCCCTCCCGGGGTGCTTTTCACCTTTCCCTCACGGTACTGGTTCACTATCGGTCACTAGGGAGTATTTAGGGTTGGGAGATGGTCCTCCCAGATTCCGACGGGATTTCACGTGTCCCGCCGTACTCAGGATACTGCTAGGTACAAAGACTATTTTAAATACGAGGCTATTACTCTCTTTGGCTGATCTTCCCAAATCATTCTTCTATAATCTTTGAGTCCACATTGCAGTCCTACAACCCCGAAGAGTAAACTCTTCGGTTTGCCCTTCTGCCGTTTCGCTCGCCGCTACTAAGGCAATCGCTTTTGCTTTCTCTTCCTGCAGCTACTTAGATGTTTCAGTTCACTGCGTCTTCCTCCTCACATCCTTAACAGATGTGGGTAACAGGTATTACCTGTTGGGTTCCCCCATTCGGAAATCCCTGGATCATCGCTTACTTACAGCTACCCAAGGTATATCGTCGTTTGTCACGTCCTTCGTCGGCTCCTAGTGCCAAGGCATCCACCGTGCGCCCTTATTAACTTAACCT"}, "NCBI_taxonomy": {"NCBI_taxonomy_name": "Streptococcus pneumoniae", "NCBI_taxonomy_id": "1313", "NCBI_taxonomy_cvterm_id": "35917"}, "protein_sequence": {"accession": "", "sequence": ""}}}}, "ARO_category": {"35950": {"category_aro_name": "antibiotic resistant gene variant or mutant", "category_aro_cvterm_id": "35950", "category_aro_accession": "0000031", "category_aro_description": "Resistance to antibiotics is often conferred by single nucleotide polymorphisms (SNPs) and other mutations in target genes."}, "36454": {"category_aro_name": "determinant of macrolide resistance", "category_aro_cvterm_id": "36454", "category_aro_accession": "3000315", "category_aro_description": "Enzymes, other proteins or other gene products shown clinically to confer resistance to macrolide antibiotics."}}, "ARO_name": "Streptococcus pneumoniae 23S rRNA with mutation conferring resistance to macrolide antibiotics", "model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "ARO_id": "41320", "model_name": "Streptococcus pneumoniae 23S rRNA with mutation conferring resistance to macrolide antibiotics", "model_type_id": "40295"}, "2816": {"model_id": "2816", "ARO_accession": "3004169", "model_param": {"blastn_bit_score": {"param_value": "5000", "param_type_id": "41093", "param_type": "BLASTN bit-score", "param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}, "snp": {"param_type": "single resistance variant", "param_value": {"7926": "A2267G", "7925": "A2266G"}, "clinical": {"7926": "A2267G", "7925": "A2266G"}, "param_type_id": "36301", "param_description": "A nucleotide or amino acid substitution that confers elevated resistance to antibiotic(s) relative to wild type. The most common type encoded in the CARD is an amino acid substitution gleaned from the literature with format [wild-type][position][mutation], e.g. R184Q. When present in the associated gene or protein, a single resistance variant confers resistance to an antibiotic drug or drug class. Single resistance variants are used by the protein variant and rRNA mutation models to detect antibiotic resistance from submitted sequences."}}, "ARO_description": "Point mutation in the 23S rRNA of Mycobaccterium smegmatis shown to confer resistance to clarithromycin, a macrolide type antibiotic.", "model_sequences": {"sequence": {"4161": {"dna_sequence": {"fmax": "3162", "fmin": "0", "accession": "AB011184.1", "strand": "+", "sequence": "TTGTAAGTGTTTAAGGGCGCATGGTGGATGCCTTGGCACTGGGAGCCGATGAAGGACGTAGGAGGCTGCGATAAGCCTCGGGGAGCTGTCAACCGAGCGTTGATCCGAGGATGTCCGAATGGGGAAACCCGGCACGAGTGATGTCGTGTCACCAGGCGCTGAATATATAGGCGTCTGGGGGGAACGCGGGGAAGTGAAACATCTCAGTACCCGTAGGAAGAGAAAACAAAATGTGATTCCGTGAGTAGTGGCGAGCGAAAGCGGAGGATGGCTAAACCGTATGCATGTGATACCGGGTAGGGGTTGTGTGTGCGGGGTTGTGGGACCTATCTTTCCGGCTCTACCTGGCTGGAGGGCAGTGAGAAAATGTTGTGGTTAGCGGAAATGGCTTGGGATGGCCTGCCGTAGACGGTGAGAGCCCGGTACGTGAAAACCCGACGTCTGTCTTGATGGTGTTCCCGAGTAGCAGCGGGCCCGTGGAATCTGCTGTGAATCTGCCGGGACCACCCGGTAAGCCTGAATACTTCCCAGTGACCGATAGCGGATTAGTACCGTGAGGGAATGGTGAAAAGTACCCCGGGAGGGGAGTGAAAGAGTACCTGAAACCGTGCGCTTACAATCCGTCAGAGCCCTCGACGTGTCGTGGGGTGATGGCGTGCCTTTTGAAGAATGAGCCTGCGAGTCAGGGACATGTCGCGAGGTTAACCCGGGTGGGGTAGCCGCAGCGAAAGCGAGTCTGAATAGGGCGTATCCACGCAACAGTGTGTGGTGTAGTGGTGTGTTCTGGACCCGAAGCGGAGTGATCTACCCATGGCCAGGGTGAAGCGCGGGTAAGACCGCGTGGAGGCCCGAACCCACTTAGGTTGAAGACTGAGGGGATGAGCTGTGGGTAGGGGTGAAAGGCCAATCAAACTCCGTGATAGCTGGTTCTCCCCGAAATGCATTTAGGTGCAGCGTTGCGTGTTTCTTGCCGGAGGTAGAGCTACTGGATGGCCGATGGGCCCCACAGGGTTACTGACGTCAGCCAAACTCCGAATGCCGGTAAGTCCAAGAGTGCGGCAGTGAGACGGCGGGGGATAAGCTCCGTGCGTCGAGAGGGAAACAGCCCAGATCGCCGGCTAAGGCCCCTAAGCGTGTGCTAAGTGGAAAAGGATGTGCAGTCGCGAAGACAACCAGGAGGTTGGCTTAGAAGCAGCCACCCTTGAAAGAGTGCGTAATAGCTCACTGGTCAAGTGATTGTGCGCCGATAATGTAGCGGGGCTCAAGCACACCGCCGAAGCCGCGGCAACGACCTTGTGTCGTTGGGTAGGGGAGCGTCCTGCATCCGGTGAAGCCGCCGAGTGATCGAGTGGTGGAGGGTGTGGGAGTGAGAATGCAGGCATGAGTAGCGATTAGGCAAGTGAGAACCTTGCCCGCCGAAAGACCAAGGGTTCCTGGGCCAGGCCAGTCCGCCCAGGGTGAGTCGGGACCTAAGGCGAGGCCGACAGGCGTAGTCGATGGACAACGGGTTGATATTCCCGTACCCGTGTATGTGCGTCCATGATGAATCAGCGGTACTAACCATCCAAAACCACCGTGACTGCACCTTTCGGGGTGTGGCGTTGGTGGGGCTGCATGGGACCTTCGTTGGTAGTAGTCAAGCGATGGGGTGACGCAGGAAGGTAGCCGTACCGGTCAGTGGTAATACCGGGGTAAGCCTGTAGGGAGTCAGATAGGTAAATCCGTCTGGCATATATCCTGAGAGGTGATGCATAGCCGAGTGAGGCGAATTCGGTGATCCTATGCTGCCGAGAAAAGCCTCTAGCGAGGACATACACGGCCCGTACCCCAAACCAACACAGGTGGTCAGGTAGAGAATACTAAGGCGTACGAGTGAACTATGGTTAAGGAACTCGGCAAAATGCCCCCGTAACTTCGGGAGAAGGGGGACCCACATGGCGTGTAAGCCTTTACGGCCCAAGCGTGAGTGGGTGGCACAAACCAGTGAGAAGCGACTGTTTACTAAAAACACAGGTCCGTGCGAAGTCGCAAGACGATGTATACGGACTGACGCCTGCCCGGTGCTGGAAGGTTAAGAGGACCCGTTAACCCTTCGGGGTGAAGCGGAGAATTTAAGCCCCAGTAAACGGCGGTGGTAACTATAACCATCCTAAGGTAGCGAAATTCCTTGTCGGGTAAGTTCCGACCTGCACGAATGGCGTAACGACTTCTCAACTGTCTCAACCATAGACTCGGCGAAATTGCACTACGAGTAAAGATGCTCGTTACGCGCGGCAGGACGAAAAGACCCCGGGACCTTCACTACAACTTGGTATTGGTGCTCGATACGGTTTGTGTAGGATAGGTGGGAGACTGTGAAGCTCACACGCCAGTGTGGGTGGAGTCGTTGTTGAAATACCACTCTGATCGTATTGGGCCTCTAACCTCGGACCGTATATCCGGTTCAGGGACAGTGCCTGGTGGGTAGTTTAACTGGGGCGGTTGCCTCCTAAAATGTAACGGAGGCGCCCAAAGGTTCCCTCAACCTGGACGGCAATCAGGTGTTGAGTGTAAGTGCACAAGGGAGCTTGACTGCGAGACGGACATGTCGAGCAGGGACGAAAGTCGGGACTAGTGATCCGGCACCTCTGAGTGGAAGGGGTGTCGCTCAACGGATAAAAGGTACCCCGGGGATAACAGGCTGATCTTCCCCAAGAGTCCATATCGACGGGATGGTTTGGCACCTCGATGTCGGCTCGTCGCATCCTGGGGCTGGAGCAGGTCCCAAGGGTTGGGCTGTTCGCCCATTAAAGCGGCACGCGAGCTGGGTTTAGAACGTCGTGAGACAGTTCGGTCTCTATCCGCCGCGCGCGTCAGAAGCTTGAGGAAACCTGTCCCTAGTACGAGAGGACCGGGACGGACGAACCTCTGGTATACCAGTTGTCCCACCAGGGGCACGGCTGGATAGCCACGTTCGGACAGGATAACCGCTGAAAGCATCTAAGCGGGAAACCTCTTCCAAGACCAGGCTTCTCACCCTCTAGGAGGGATAAGGCCCCCCGCAGACCACGGGATTGATAGACCAGACCTGGAAGCCTAGTAATAGGTGCAGGGAACTGGCACTAACCGGCCGAAAACTTACAACACCCCATAATCGTTGTAAGAAGAAAACATTGACGCACC"}, "NCBI_taxonomy": {"NCBI_taxonomy_name": "Mycobacterium smegmatis", "NCBI_taxonomy_id": "1772", "NCBI_taxonomy_cvterm_id": "36871"}, "protein_sequence": {"accession": "", "sequence": ""}}}}, "ARO_category": {"36454": {"category_aro_name": "determinant of macrolide resistance", "category_aro_cvterm_id": "36454", "category_aro_accession": "3000315", "category_aro_description": "Enzymes, other proteins or other gene products shown clinically to confer resistance to macrolide antibiotics."}, "35950": {"category_aro_name": "antibiotic resistant gene variant or mutant", "category_aro_cvterm_id": "35950", "category_aro_accession": "0000031", "category_aro_description": "Resistance to antibiotics is often conferred by single nucleotide polymorphisms (SNPs) and other mutations in target genes."}}, "ARO_name": "Mycobacterium smegmatis 23S rRNA with mutation conferring resistance to clarithromycin", "model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "ARO_id": "41319", "model_name": "Mycobacterium smegmatis 23S rRNA with mutation conferring resistance to clarithromycin", "model_type_id": "40295"}, "2811": {"model_id": "2811", "ARO_accession": "3004164", "model_param": {"blastn_bit_score": {"param_value": "5700", "param_type_id": "41093", "param_type": "BLASTN bit-score", "param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}, "snp": {"param_type": "single resistance variant", "param_value": {"7913": "A2272C", "7915": "A2274U", "7914": "A2274G"}, "clinical": {"7913": "A2272C", "7915": "A2274U", "7914": "A2274G"}, "param_type_id": "36301", "param_description": "A nucleotide or amino acid substitution that confers elevated resistance to antibiotic(s) relative to wild type. The most common type encoded in the CARD is an amino acid substitution gleaned from the literature with format [wild-type][position][mutation], e.g. R184Q. When present in the associated gene or protein, a single resistance variant confers resistance to an antibiotic drug or drug class. Single resistance variants are used by the protein variant and rRNA mutation models to detect antibiotic resistance from submitted sequences."}}, "ARO_description": "Point mutation in the 23S rRNA of Mycobacterium avium shown to confer resistance to clarithromycin, a macrolide type antibiotic.", "model_sequences": {"sequence": {"4157": {"dna_sequence": {"fmax": "3112", "fmin": "0", "accession": "NG_041979.1", "strand": "+", "sequence": "TGTGTGTAAGTAAGTGTTTAAGGGCGCATGGTGGATGCCTTGGCATCGAGAGCCGATGAAGGACGTGGGAGGCTGCGATATGCCTCGGGGAGCTGTCAACCGAGCATTGATCCGAGGATTTCCGAATGGGGAAACCCAGCACGAGTGATGTCGTGTTACCCGTATCTGAATATATAGGGTGCGGGAGGTAACGCGGGGAAGTGAAACATCTCAGTACCCGTAGGAGAAGAAAACAATTGTGATTCCGTCAGTAGTGGCGAGCGAACGCGGAACAGGCTAAACCGCATGCATGGACAACCGGGTAGGGGTTGTGTGTGCGGGGTTGTGGGATTGATATGTCTCAGCTCTACCTGGCTGAGGGGTAGTCAGAAAGTGTCGTGGTTAGCGGAAGTGGCCTGGGACGGCCCGCCGTAGACGGTGAGAGCCCGGTACGCGAAAACCCGGCACCTGCCTTATATCAACACCCGAGTAGCAGCGGGCCCGTGGAATCTGCTGTGAATCTGCCGGGACCACCCGGTAAGCCTAAATACTTCTCGATGACCGATAGCGGATTAGTACCGTGAGGGAATGGTGAAAAGTACCCCGGGAGGGGAGTGAAATAGTACCTGAAACCGTGTGCCTACAATCCGTCAGAGCCTCCTCGTGGGGTGATGGCGTGCCTTTTGAAGAATGAGCCTGCGAGTCAGGGACACGTCGCGAGGTTAACCCGTGCGGGGTAGCCGCAGCGAAAGCGAGTCTGAATAGGGCGCATCCCCTTTGGGGTGTAGTGGCGTGTTCTGGACCCGAAGCGGAGTGATCTACCCATGGCCAGGGTGAAGCGCGGGTAAGACCGCGTGGAGGCCCGAACCCACTTAGGTTGAAGACTGAGGGGATGAGCTGTGGGTAGGGGTGAAAGGCCAATCAAACTCCGTGATAGCTGGTTCTCCCCGAAATGCATTTAGGTGCAGCGTTGCGTGGTTCACCACGGAGGTAGAGCTACTGGATGGCCGATGGGCCCTACTAGGTTACTGACGTCAGCCAAACTCCGAATGCCGTGGTGTAAAGCGTGGCAGTGAGACGGCGGGGGATAAGCTCCGTACGTCGAAAGGGAAACAGCCCAGATCGCCGGCTAAGGCCCCTAAGCGTGTGCTAAGTGGAAAAGGATGTGTAGTCGCAGAGACAACCAGGAGGTTGGCTTAGAAGCAGCCACCCTTGAAAGAGTGCGTAATAGCTCACTGGTCAAGTGATTATGCGCCGATAATGTAGCGGGGCTCAAGCACACCGCCGAAGCCGCGGCACATTCATCTTTACGGTGGATGTGGGTAGGGGAGCGTCCCCCATTCAGCGAAGCCTCCGGGTGACCGGTGGTGGAGGGTGGGGGAGTGAGAATGCAGGCATGAGTAGCGATAAGGCAAGTGAGAACCTTGCCCGCCGTAAGACCAAGGGTTCCTGGGCCAGGCCAGTCCGCCCAGGGTGAGTCGGGACCTAAGGCGAGGCCGACAGGCGTAGTCGATGGACAACGGGTTGATATTCCCGTACCCGTGTATGGGCGTCCCTGATGAATCAGCGGTACTAACCACCCAAAACCGGATCGACCATTCCCCTTCGGGGGCGTGGCGATTCGGGGCTGCGTGGGACCTTCGCTGGTAGTAGTCAAGCAATGGGGTGACGCAGGAAGGCAGCCGTACCAGTCAGTGGTAATACTGGGGCAAGCCCGTAGGGAGAGCGATAGGCAAATCCGTCGCTCACTAATCCTGAGAGGTGATGCATAGCCGGTTGAGGCGAATTCGGTGATCCTCTGCTGCCAAGAAAAGCCTCTAGCGAGCACATACACGGCCCGTACCCCAAACCAACACAGGTGGTCAGGTAGAGAATACCAAGGCGTACGAGATAACTATGGTTAAGGAACTCGGCAAAATGCCCCCGTAACTTCGGGAGAAGGGGGGCCGGAATACCGTGAACACCCTTGCGGTGGGAGCGGGATTCGGCCGCAGAAACCAGTGGGTAGCGACTGTTTACTAAAAACACAGGTCCGTGCGAAGTCGCAAGACGATGTATACGGACTGACGCCTGCCCGGTGCTGGAAGGTTAAGAGGACCCGTTAACCCGTAAGGGTGAAGCGGAGAATTTAAGCCCCAGTAAACGGCGGTGGTAACTATAACCATCCTAAGGTAGCGAAATTCCTTGTCGGGTAAGTTCCGACCTGCACGAATGGCGTAACGACTTCCCAACTGTCTCAACCATAGACTCGGCGAAATTGCACTACGAGTAAAGATGCTCGTTACGCGCGGCAGGACGAAAAGACCCCGGGACCTTCACTACAACTTGGTATTGGTGTTCGGTACGGTTTGTGTAGGATAGGTGGGAGACTTTGAAGCACAGACGCCAGTTTGTGTGGAGTCGTTGTTGAAATACCACTCTGATCGTATTGGACACCTAACGTCGAACCCTTATCGGGTTCACGGACAGTGCCTGGCGGGTAGTTTAACTGGGGCGGTTGCCTCCTAAAATGTAACGGAGGCGCCCAAAGGTTCCCTCAACCTGGACGGCAATCAGGTGGCGAGTGTAAGTGCACAAGGGAGCTTGACTGCGAGACTTACAAGTCAAGCAGGGACGAAAGTCGGGACTAGTGATCCGGCACCCCCGAGTGGAAGGGGTGTCGCTCAACGGATAAAAGGTACCCCGGGGATAACAGGCTGATCTTCCCCAAGAGTCCATATCGACGGGATGGTTTGGCACCTCGATGTCGGCTCGTCGCATCCTGGGGCTGGAGCAGGTCCCAAGGGTTGGGCTGTTCGCCCATTAAAGCGGCACGCGAGCTGGGTTTAGAACGTCGTGAGACAGTTCGGTCTCTATCCGCCGCGCGCGTCAGAAACTTGAGGAAACCTGTCCCTAGTACGAGAGGACCGGGACGGACGAACCTCTGGTATACCAGTTGTCCCACCAGGGGCACGGCTGGATAGCCACGTTCGGACAGGATAACCGCTGAAAGCATCTAAGCGGGAAACCTTCTCCAAGATCAGGTTTCTCACCCTTTTAGAGGGATAAGGCCCCCCGCAGACCACGGGATTGATAGGCCAGACCTGGAAGCTCAGTAATGAGTGCAGGGAACTGGCACTAACCGGCCGAAAAC"}, "NCBI_taxonomy": {"NCBI_taxonomy_name": "Mycobacterium avium", "NCBI_taxonomy_id": "1764", "NCBI_taxonomy_cvterm_id": "39582"}, "protein_sequence": {"accession": "", "sequence": ""}}}}, "ARO_category": {"35950": {"category_aro_name": "antibiotic resistant gene variant or mutant", "category_aro_cvterm_id": "35950", "category_aro_accession": "0000031", "category_aro_description": "Resistance to antibiotics is often conferred by single nucleotide polymorphisms (SNPs) and other mutations in target genes."}, "36454": {"category_aro_name": "determinant of macrolide resistance", "category_aro_cvterm_id": "36454", "category_aro_accession": "3000315", "category_aro_description": "Enzymes, other proteins or other gene products shown clinically to confer resistance to macrolide antibiotics."}}, "ARO_name": "Mycobacterium avium 23S rRNA with mutation conferring resistance to clarithromycin", "model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "ARO_id": "41312", "model_name": "Mycobacterium avium 23S rRNA with mutation conferring resistance to clarithromycin", "model_type_id": "40295"}, "2810": {"model_id": "2810", "ARO_accession": "3004163", "model_param": {"blastn_bit_score": {"param_value": "5700", "param_type_id": "41093", "param_type": "BLASTN bit-score", "param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}, "snp": {"param_type": "single resistance variant", "param_value": {"7912": "A2272C", "7911": "A2272G", "7910": "A2271G"}, "clinical": {"7912": "A2272C", "7911": "A2272G", "7910": "A2271G"}, "param_type_id": "36301", "param_description": "A nucleotide or amino acid substitution that confers elevated resistance to antibiotic(s) relative to wild type. The most common type encoded in the CARD is an amino acid substitution gleaned from the literature with format [wild-type][position][mutation], e.g. R184Q. When present in the associated gene or protein, a single resistance variant confers resistance to an antibiotic drug or drug class. Single resistance variants are used by the protein variant and rRNA mutation models to detect antibiotic resistance from submitted sequences."}}, "ARO_description": "Point mutation in the 23S rRNA of Mycobacterium abscessus shown to confer resistance to clarithromycin, a macrolide type antibiotic.", "model_sequences": {"sequence": {"4156": {"dna_sequence": {"fmax": "1467319", "fmin": "1464207", "accession": "NC_010397.1", "strand": "+", "sequence": "CTAAGTTCTTAAGGGCGCATGGTGAATGCCTTGGCACTAGAAGCCGAAGAAGGACGTAGGAGGCTGCGATAAGCCTCGGGGAGCTGCCAACCGAGCTTTGATCCGAGGATCTCCGAATGGGGAAACCCAGCACGAGTGATGTCGTGTTACCCACTGCTGAATATATAGGCTTTGGGAGGAAACGCGGGGAAGTGAAACATCTCAGTACCCGTAGGAAGAGAAAACAACCGTGATTCCGTGAGTAGTGGCGAGCGAAAGCGGAAGATGGCTAAACCGCATGCATGTGATACCTGGTAGGGGTTGTGTGTGCGGGGTTGTGGGAGTTGTACTTGCCGGTTCTACCAGGCCGGCGGACAGTAAAAAAGTGTCGTGATTAGCGGAAGTGGTCTGGGACGGCCCGCCGCAGACGGTGAGAGTCCGGTACGCGAAAATCCGACACCTGTCTCGTACTTCATCCCGAGTAGCAGCGGGCTCGTGGAATCTGCTGTGAATCTGCCGGGACCACCCGGTAAGCCTAAATACTCTCTAGTGACCGATAGCGGATTAGTACCGTGAGGGAATGGTGAAAAGTACCCCGGGAGGGGAGTGAAATAGTACCTGAAACCATGTGCCTACAATCCGTCAAAGCCTCCTCGTGGGGTGATGGCGTGCCTTTTGAAGAATGAGCCTGCGAGTCAGGGACACGTCGCGAGGTTAACCCGTGTGGGGTAGCCGTAGCGAAAGCGAGTCTGAATAGGGCGCCCATAGTGGCGTGTTCTGGACCCGAAGCGGAGTGATCTACCCATGGCCAGGGTGAAGCGGCGGTAAGACGCCGTGGAGGCCCGAACCCACTTAGGTTGAAGACTGAGGGGATGAGCTGTGGGTAGGGGTGAAAGGCCAATCAAACTCCGTGATAGCTGGTTCTCCCCGAAATGCATTTAGGTGCAGCGTCGCGTGTTTCTTGTTGGAGGTAGAGCTACTGGATGGCCGATGGGCCCTACTAGGTTACTGACGTCAGCCAAACTCCGAATGCCAATAAGTTAGAGCGCGGCAGTGAGACGGCGGGGGAGAAGCTCCGTACGTCGAGAGGGAAACAGCCCAGATCGCCGGCTAAGGCCCCTAAGCGTGTACTAAGTGGAAAAGGATGTGCAGTCGCGAAGACAACCAGGAGGTTGGCTTAGAAGCAGCCACCCTTGAAAGAGTGCGTAATAGCTCACTGGTCAAGTGATTGTGCGCCGACAATGTAGCGGGGCTCAAGTACACCGCCGAAGCCGCGGCATTCATGCAATACATTCCCTTCGGGGCAGTGGCATGGATGGGTAGGGGAGCGTCCTGCACCCAGCGAAGCTGCGAAGTAATTCAGCAGTGGAGGGTGCGGGAGTGAGAATGCAGGCATGAGTAGCGACAGGCAAGTGAGAAACTTGCCCGCCGAATGACCAAGGGTTCCTGGGCCAGGCTAGTCCTCCCAGGGTAAGTCGGGACCTAAGGCGAGGCCGACAGGCGTAGTCGATGGACAACGGGTTGATATTCCCGTACCCGTGTGTGCGCGCCCATGATGAATCATCGGTACTAACCACCCAAAAGGTTCTAGATCAATCTCTTCGGAGTGCGACGTGAACCCGCTGCGTGGGACCTTCGGTGGTAGTAGTCAAGCGATGGGGTGACGCAGGAAGGTAGCTGTACCGGTTAGTGGTTATACCGGAGCAAGCCCGTAGGACGACATCTAGGCAAATCCGGATGTCATACGTCTGAGAGGTGACGCATAGCCGATTGAGGCGAATTCAGTGATCCTATGCTGCCAAGAAAAGCCTCTAGTGAGTTCACACACGGCCCGTACCCCAAACCAACACAGGTGGTCAGGTAGAGAATACTAAGGCGTACGAGATAACTATGGTTAAGGAACTCGGCAAAATACCCCCGTAACTTCGGGAGAAGGGGGACCTCGCTTGGTGACCGGACTTGCTCCGTGAGCTGAACGAGGTCGCAGAGACCAGTGAGAAGCGACTGTTTACTAAAAACACAGGTCCGTGCGAAGTCGCAAGACGATGTATACGGACTGACGCCTGCCCGGTGCTGGAAGGTTAAGAGGACCCGTTAACCCTTGGGTGAAGCGGAGAATTTAAGCCCCAGTAAACGGCGGTGGTAACTATAACCATCCTAAGGTAGCGAAATTCCTTGTCGGGTAAGTTCCGACCTGCACGAATGGCGTAACGACTTCTCAACTGTCTCAACCATAGACTCGGCGAAATTGCACTACGAGTAAAGATGCTCGTTACGCGCGGCAGGACGAAAAGACCCCGGGACCTTCACTATAGCTTGGTATTGGCGTTTGGTTCGGTTTGTGTAGGATAGGTGGGAGACTGTGAAGCAGGCACGCCAGTGTTTGTGGAGTCATCGTTGAAATACCACTCTGATCGTATTGAACCTCTAACCTCGGACCGTATATCCGGTCCAGGGACAGTGCCTGGTGGGTAGTTTAACTGGGGCGGTTGCCTCCCAAAATGTAACGGAGGCGCCCAAAGGTTCCCTCAACCTGGACGGCAATCAGGTGTTGAGTGCAAGTGCACAAGGGAGCTTGACTGCGAGACCTACAAGTCAAGCAGGGACGAAAGTCGGGACTAGTGATCCGGCACCTCTGAGTGGAAGGGGTGTCGCTCAACGGATAAAAGGTACCCCGGGGATAACAGGCTGATCTTCCCCAAGAGTCCATATCGACGGGATGGTTTGGCACCTCGATGTCGGCTCGTCGCATCCTGGGGCTGGAGCAGGTCCCAAGGGTTGGGCTGTTCGCCCATTAAAGCGGCACGCGAGCTGGGTTTAGAACGTCGTGAGACAGTTCGGTCTCTATCCGCCGCGCGCGTCAGAAACTTGAGGAAACCTGTCCCTAGTACGAGAGGACCGGGACGGACGAACCTCTGGTGTACCAGTTGTTCCACCAGGAGCACGGCTGGATAGCTACGTTCGGACAGGATAACCGCTGAAAGCATCTAAGCGGGAAACCTATTCCAAGACCAGGTTTCTTACCCTTTTAGAGGGATAAGGTCACCCACAGACTATGGGTTCAATAGGCCAGACCTGCAAGCGTAGTAATACGTTCAGGGAACTGGCACTAATCGACCGAAAACTTACTAAT"}, "NCBI_taxonomy": {"NCBI_taxonomy_name": "Mycobacterium abscessus", "NCBI_taxonomy_id": "36809", "NCBI_taxonomy_cvterm_id": "36888"}, "protein_sequence": {"accession": "", "sequence": ""}}}}, "ARO_category": {"35950": {"category_aro_name": "antibiotic resistant gene variant or mutant", "category_aro_cvterm_id": "35950", "category_aro_accession": "0000031", "category_aro_description": "Resistance to antibiotics is often conferred by single nucleotide polymorphisms (SNPs) and other mutations in target genes."}, "36454": {"category_aro_name": "determinant of macrolide resistance", "category_aro_cvterm_id": "36454", "category_aro_accession": "3000315", "category_aro_description": "Enzymes, other proteins or other gene products shown clinically to confer resistance to macrolide antibiotics."}}, "ARO_name": "Mycobacterium abscessus 23S rRNA with mutation conferring resistance to clarithromycin", "model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "ARO_id": "41311", "model_name": "Mycobaccterium abscessus 23S rRNA with mutation conferring resistance to clarithromycin", "model_type_id": "40295"}, "2813": {"model_id": "2813", "ARO_accession": "3004166", "model_param": {"blastn_bit_score": {"param_value": "5700", "param_type_id": "41093", "param_type": "BLASTN bit-score", "param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}, "snp": {"param_type": "single resistance variant", "param_value": {"8172": "A2266U", "8173": "A2266C", "8174": "A2266G"}, "clinical": {"8172": "A2266U", "8173": "A2266C", "8174": "A2266G"}, "param_type_id": "36301", "param_description": "A nucleotide or amino acid substitution that confers elevated resistance to antibiotic(s) relative to wild type. The most common type encoded in the CARD is an amino acid substitution gleaned from the literature with format [wild-type][position][mutation], e.g. R184Q. When present in the associated gene or protein, a single resistance variant confers resistance to an antibiotic drug or drug class. Single resistance variants are used by the protein variant and rRNA mutation models to detect antibiotic resistance from submitted sequences."}}, "ARO_description": "Point mutation in the 23S rRNA of Mycobacterium intracellulare shown to confer resistance to clarithromycin, a macrolide type antibiotic.", "model_sequences": {"sequence": {"4177": {"dna_sequence": {"fmax": "3103", "fmin": "0", "accession": "NR_076151.1", "strand": "+", "sequence": "TAAGTGTTTAAGGGCGCATGGTGGATGCCTTGGCATCGAGAGCCGATGAAGGACGTGGGAGGCTGCGATATGCCTCGGGGAGCTGTCAACCGAGCATTGATCCGAGGATTTCCGAATGGGGAAACCCAGCACGAGTGATGTCGTGTTACCCGCATCTGAATATATAGGGTGCGGGAGGGAACGCGGGGAAGTGAAACATCTCAGTACCCGTAGGAGAAGAAAACAATTGTGATTCCGTAAGTAGTGGCGAGCGAACGCGGAACAGGCTAAACCGCACGCATGTGATACCGGGTAGGGGTTGTGTGTGCGGGGTTGTGGGAGGATACATCTCAGCTCTACCTGGCTGAGGGGTAGTCAGAAAGTGTCGTGGTTAGCGGAAGTGGCCTGGGATGGTCTGCCGTAGACGGTGAGAGCCCGGTACGCGAAAACCCGTCACCTACCTTGTATCAATTCCCGAGTAGCAGCGGGCCCGTGGAATCTGCTGTGAATCTGCCGGGACCACCCGGTAAGCCTAAATACTTCTCGATGACCGATAGCGGATTAGTACCGTGAGGGAATGGTGAAAAGTACCCCGGGAGGGGAGTGAAATAGTACCTGAAACCGTGTGCCTACAATCCGTCAGAGCCTCCTTGTGGGGTGATGGCGTGCCTTTTGAAGAATGAGCCTGCGAGTCAGGGACACGTCGCGAGGTTAACCCGTGCGGGGTAGCCGCAGCGAAAGCGAGTCTGAATAGGGCGCATCCCCTTTGGGGTGTAGTGGCGTGTTCTGGACCCGAAGCGGAGTGATCTACCCATGGCCAGGGTGAAGCGCGGGTAAGACCGCGTGGAGGCCCGAACCCACTTAGGTTGAAGACTGAGGGGATGAGCTGTGGGTAGGGGTGAAAGGCCAATCAAACTCCGTGATAGCTGGTTCTCCCCGAAATGCATTTAGGTGCAGCGTTGCGTGGTTCACCACGGAGGTAGAGCTACTGGATGGCCGATGGGCCCTACTAGGTTACTGACGTCAGCCAAACTCCGAATGCCGTGGTGTAAAGCGTGGCAGTGAGACGGCGGGGGATAAGCTCCGTACGTCGAAAGGGAAACAGCCCAGATCGCCGGCTAAGGCCCCTAAGCGTGTGCTAAGTGGAAAAGGATGTGTAGTCGCAGAGACAACCAGGAGGTTGGCTTAGAAGCAGCCACCCTTGAAAGAGTGCGTAATAGCTCACTGGTCAAGTGATTATGCGCCGATAATGTAGCGGGGCTCAAGCACACCGCCGAAGCCGCGGCACATTCACGTTTACGTGGATGTGGGTAGGGGGAGCGTCCCTCATTCAGCGAAGCCTCCGGGTGACCGGTGGTGGAGGGTGGGGGAGTGAGAATGCAGGCATGAGTAGCGATAAGGCAAGTGAGAACCTTGCCCGCCGTAAGACCAAGGGTTCCTGGGCCAGGCCAGTCCGCCCAGGGTGAGTCGGGACCTAAGGCGAGGCCGACAGGCGTAGTCGATGGACAACGGGTTGATATTCCCGTACCCGTGTATGGGCGTCCCTGATGAATCAGCGGTACTAACCACCCAAAACCGGATCGACCATTCCCCTTCGGGGGCATGGAGTTTCGGGGCTGCGTGGGACCTTCGCTGGTAGTAGTCAAGCAATGGGGTGACGCAGGAAGGTAGCCGTACCAGTCAGTGGTAATACTGGGGCAAGCCTGTAGGGAGAGCGATAGGCAAATCCGTCGCTCATTAATCCTGAGAGGTGATGCATAGCCGATTGAGGTGAATTCGGTGATCCTCTGCTGCCAAGAAAAGCCTCTAGCGAGCACATACACGGCCCGTACCCCAAACCAACACAGGTGGTCAGGTAGAGAATACCAAGGCGTACGAGATAACTATGGTTAAGGAACTCGGCAAAATGCCCCCGTAACTTCGGGAGAAGGGGGGCCGGAATACCGTGAACACCCTTGCGGTGGGAGCGGGATCCGGCCGCAGAAACCAGTGGGTAGCGACTGTTTACTAAAAACACAGGTCCGTGCGAAGTCGCAAGACGATGTATACGGACTGACGCCTGCCCGGTGCTGGAAGGTTAAGAGGACCCGTTAACCGTAAGGTGAAGCGGAGAATTTAAGCCCCAGTAAACGGCGGTGGTAACTATAACCATCCTAAGGTAGCGAAATTCCTTGTCGGGTAAGTTCCGACCTGCACGAATGGCGTAACGACTTCCCAACTGTCTCAACCATAGACTCGGCGAAATTGCACTACGAGTAAAGATGCTCGTTACGCGCGGCAGGACGAAAAGACCCCGGGACCTTCACTACAACTTGGTATTGGTGTTCGGTACGGTTTGTGTAGGATAGGTGGGAGACTGTGAAATACAGACGCCAGTTTGTATGGAGTCGTTGTTGAAATACCACTCTGATCGTATTGGACACCTAACGTCGAACCCTTATCGGGTTCACGGACAGTGCCTGGCGGGTAGTTTAACTGGGGCGGTTGCCTCCTAAAATGTAACGGAGGCGCCCAAAGGTTCCCTCAACCTGGACGGCAATCAGGTGACGAGTGTAAGTGCACAAGGGAGCTTGACTGCGAGACTTACAAGTCAAGCAGGGACGAAAGTCGGGACTAGTGATCCGGCACCCCCGAGTGGAAGGGGTGTCGCTCAACGGATAAAAGGTACCCCGGGGATAACAGGCTGATCTTCCCCAAGAGTCCATATCGACGGGATGGTTTGGCACCTCGATGTCGGCTCGTCGCATCCTGGGGCTGGAGCAGGTCCCAAGGGTTGGGCTGTTCGCCCATTAAAGCGGCACGCGAGCTGGGTTTAGAACGTCGTGAGACAGTTCGGTCTCTATCCGCCGCGCGCGTCAGAAACTTGAGGAAACCTGTCCCTAGTACGAGAGGACCGGGACGGACGAACCTCTGGTATACCAGTTGTTCCACCAGGAGCACGGCTGGATAGCCACGTTCGGACAGGATAACCGCTGAAAGCATCTAAGCGGGAAACCTTCTCCAAGATCAGGTTTCTCACCCTTTTAGAGGGATAAGGCCCCCCGCAGACCACGGGTTCGATAGGCTAGACCTGGAAGCTCAGCAATGAGTGCAGGGAACTGGCACTAACCGGCCGAAAACTTAC"}, "NCBI_taxonomy": {"NCBI_taxonomy_name": "Mycobacterium intracellulare", "NCBI_taxonomy_id": "1767", "NCBI_taxonomy_cvterm_id": "41317"}, "protein_sequence": {"accession": "", "sequence": ""}}}}, "ARO_category": {"35950": {"category_aro_name": "antibiotic resistant gene variant or mutant", "category_aro_cvterm_id": "35950", "category_aro_accession": "0000031", "category_aro_description": "Resistance to antibiotics is often conferred by single nucleotide polymorphisms (SNPs) and other mutations in target genes."}, "36454": {"category_aro_name": "determinant of macrolide resistance", "category_aro_cvterm_id": "36454", "category_aro_accession": "3000315", "category_aro_description": "Enzymes, other proteins or other gene products shown clinically to confer resistance to macrolide antibiotics."}}, "ARO_name": "Mycobacterium intracellulare 23S rRNA with mutation conferring resistance to clarithromycin", "model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "ARO_id": "41315", "model_name": "Mycobacterium intracellulare 23S rRNA with mutation conferring resistance to clarithromycin", "model_type_id": "40295"}, "2812": {"model_id": "2812", "ARO_accession": "3004165", "model_param": {"blastn_bit_score": {"param_value": "5700", "param_type_id": "41093", "param_type": "BLASTN bit-score", "param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}, "snp": {"param_type": "single resistance variant", "param_value": {"7919": "A2273G", "7918": "A2273C", "7917": "A2272G", "7916": "A2272C"}, "clinical": {"7919": "A2273G", "7918": "A2273C", "7917": "A2272G", "7916": "A2272C"}, "param_type_id": "36301", "param_description": "A nucleotide or amino acid substitution that confers elevated resistance to antibiotic(s) relative to wild type. The most common type encoded in the CARD is an amino acid substitution gleaned from the literature with format [wild-type][position][mutation], e.g. R184Q. When present in the associated gene or protein, a single resistance variant confers resistance to an antibiotic drug or drug class. Single resistance variants are used by the protein variant and rRNA mutation models to detect antibiotic resistance from submitted sequences."}}, "ARO_description": "Point mutation in the 23S rRNA of Mycobacterium chelonae shown to confer resistance to clarithromycin, a macrolide type antibiotic.", "model_sequences": {"sequence": {"4158": {"dna_sequence": {"fmax": "3113", "fmin": "0", "accession": "GU143889.1", "strand": "+", "sequence": "CTAAGTTCTTAAGGGCACATGGTGAATGCCTTGGCACTAGAAGCCGAAGAAGGACGTAGGAGGCTGCGATAAGCCTCGGGGAGCTGCCAACCGAGCTTTGATCCGAGGATGTCCGAATGGGGAAACCCAGCACGAGTGATGTCGTGTTACCCACTGCTGAATATATAGGCTTTGGGAGGAAACGCGGGGAAGTGAAACATCTCAGTACCCGTAGGAAGAGAAAACAACCGTGATTCCGTGAGTAGTGGCGAGCGAAAGCGGAAGATGGCTAAACCGCATGCATGTGATACCTGGTAGGGGTTGTGTGTGCGGGGTTGTGGGAGTTGTACTTGCCGGTTCTACCAGGCCGGCGGACAGTAAAAAAGTGTCGTGATTAGCGGAAGTGGTCTGGGACGGCCCGCCGCAGACGGTGAGAGTCCGGTACGCGAAAATCCGACACCTGTCTCGTACTTCATCCCGAGTAGCAGCGGGCTCGTGGAATCTGCTGTGAATCTGCCGGGACCACCCGGTAAGCCTAAATACTCTCTAGTGACCGATAGCGGATTAGTACCGTGAGGGAATGGTGAAAAGTACCCCGGGAGGGGAGTGAAATAGTACCTGAAACCATGTGCCTACAATCCGTCAAAGCCTCCTTGTGGGGTGATGGCGTGCCTTTTGAAGAATGAGCCTGCGAGTCAGGGACACGTCGCGAGGTTAACCCGTGAGGGGTAGCCGTAGCGAAAGCGAGTCTGAATAGGGCGCCCATAGTGGCGTGTTCTGGACCCGAAGCGGAGTGATCTACCCATGGCCAGGGTGAAGCGGCGGTAAGACGCCGTGGAGGCCCGAACCCACTTAGGTTGAAGACTGAGGGGATGAGCTGTGGGTAGGGGTGAAAGGCCAATCAAACTCCGTGATAGCTGGTTCTCCCCGAAATGCATTTAGGTGCAGCGTCGCGTGTTTCTTATTGGAGGTAGAGCTACTGGATGGCCGATGGGCCCTACTAGGTTACTGACGTCAGCCAAACTCCGAATGCCAATAAGTTAGAGCGCGGCAGTGAGACGGCGGGGGAGAAGCTCCGTACGTCGAGAGGGAAACAGCCCAGATCGCCGGCTAAGGCCCCTAAGCGTGTACTAAGTGGAAAAGGATGTGCAGTCGCAAAGACAACCAGGAGGTTGGCTTAGAAGCAGCCACCCTTGAAAGAGTGCGTAATAGCTCACTGGTCAAGTGATTGTGCGCCGACAATGTAGCGGGGCTCAAGTACACCGCCGAAGCCGCGGCATTCATGCAATACATTCCCTTCGGGGCAGTGGCATGGATGGGTAGGGGAGCGTCCTGCACCCAGCGAAGCTGCGGAGTAATCCAGCAGTGGAGGGTGCGGGAGTGAGAATGCAGGCATGAGTAGCGACAGGCAAGTGAGAAACTTGCCCGCCGAATGACCAAGGGTTCCTGGGCCAGGCTAGTCCTCCCAGGGTAAGTCGGGACCTAAGGCGAGGCCGACAGGCGTAGTCGATGGACAACGGGTTGATATTCCCGTACCCGTGTGTGCGCGCCCATGATGAATCATCGGTACTAACCACCCAAAAGGTTCTAGATCAATCTCTTCGGAGTGCGACGTGAACCCGCTGCGTGGGACCTTCGGTGGTAGTAGTCAAGCGATGGGGTGACGCAGGAAGGTAGCTGTACCGGTTAGTGGTTATACCGGAGCAAGCCCGTAGGGCGACGTCTAGGTAAATCCGGATGTCATTAAGCCTGAGAGGTGACGCATAGCCGATTGAGGCGAATTCAGTGATCCTATGCTGCCAAGAAAAGCCTCTAGTGAGTTCACACACGGCCCGTACCCCAAACCAACACAGGTGGTCAGGTAGAGAATACTAAGGCGTACGAGATAACTATGGTTAAGGAACTCGGCAAAATACCCCCGTAACTTCGGGAGAAGGGGGACCTCGCTTGGTGACCGGACTTGCTCCGTGAGCTGAACGAGGTCGCAGAGACCAGTGAGAAGCGACTGTTTACTAAAAACACAGGTCCGTGCGAAGTCGCAAGACGATGTATACGGACTGACGCCTGCCCGGTGCTGGAAGGTTAAGAGGACCCGTTAACCCTTGGGTGAAGCGGAGAATTTAAGCCCCAGTAAACGGCGGTGGTAACTATAACCATCCTAAGGTAGCGAAATTCCTTGTCGGGTAAGTTCCGACCTGCACGAATGGCGTAACGACTTCTCAACTGTCTCAACCATAGACTCGGCGAAATTGCACTACGAGTAAAGATGCTCGTTACGCGCGGCAGGACGAAAAGACCCCGGGACCTTCACTATAGCTTGGTATTGGCGTTTGGTTCGGTTTGTGTAGGATAGGTGGGAGACTGTGAAGCAGGCACGCCAGTGTTTGTGGAGTCATCGTTGAAATACCACTCTGATCGTATTGAACCTCTAACCTCGGACCGTATATCCGGTCCAGGGACAGTGCCTGGTGGGTAGTTTAACTGGGGCGGTTGCCTCCCAAAATGTAACGGAGGCGCCCAAAGGTTCCCTCAACCTGGACGGCAATCAGGTGTTGAGTGCAAGTGCACAAGGGAGCTTGACTGCGAGACTTACAAGTCGAGCAGGGACGAAAGTCGGGACTAGTGATCCGGCATCTCTGAGTGGAAGGGATGTCGCTCAACGGATAAAAGGTACCCCGGGGATAACAGGCTGATCTTCCCCAAGAGTCCATATCGACGGGATGGTTTGGCACCTCGATGTCGGCTCGTCGCATCCTGGGGCTGGAGCAGGTCCCAAGGGTTGGGCTGTTCGCCCATTAAAGCGGCACGCGAGCTGGGTTTAGAACGTCGTGAGACAGTTCGGTCTCTATCCGCCGCGCGCGTCAGAAACTTGAGGAAACCTGTCCCTAGTACGAGAGGACCGGGACGGACGAACCTCTGGTGTACCAGTTGTTCCACCAGGAGCACGGCTGGATAGCTACGTTCGGACAGGATAACCGCTGAAAGCATCTAAGCGGGAAACCTATTCCAAGACCAGGTTTCTTACCCTTTTAGAGGGATAAGGTCACCCACAGACTATGGGTTCAATAGGCCAGACCTGTAAGCGTAGTAATACGTTCAGGGAACTGGCACTAATCGACCGAAAACTTACTAAT"}, "NCBI_taxonomy": {"NCBI_taxonomy_name": "Mycobacterium chelonae", "NCBI_taxonomy_id": "1774", "NCBI_taxonomy_cvterm_id": "41314"}, "protein_sequence": {"accession": "", "sequence": ""}}}}, "ARO_category": {"35950": {"category_aro_name": "antibiotic resistant gene variant or mutant", "category_aro_cvterm_id": "35950", "category_aro_accession": "0000031", "category_aro_description": "Resistance to antibiotics is often conferred by single nucleotide polymorphisms (SNPs) and other mutations in target genes."}, "36454": {"category_aro_name": "determinant of macrolide resistance", "category_aro_cvterm_id": "36454", "category_aro_accession": "3000315", "category_aro_description": "Enzymes, other proteins or other gene products shown clinically to confer resistance to macrolide antibiotics."}}, "ARO_name": "Mycobacterium chelonae 23S rRNA with mutation conferring resistance to clarithromycin", "model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "ARO_id": "41313", "model_name": "Mycobacterium chelonae 23S rRNA with mutation conferring resistance to clarithromycin", "model_type_id": "40295"}, "2819": {"model_id": "2819", "ARO_accession": "3004173", "model_param": {"blastn_bit_score": {"param_value": "5000", "param_type_id": "41093", "param_type": "BLASTN bit-score", "param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}, "snp": {"param_type": "single resistance variant", "param_value": {"7933": "G2032U", "7932": "G2032A", "7935": "G2447U", "7934": "G2032C"}, "clinical": {"7933": "G2032U", "7932": "G2032A", "7935": "G2447U", "7934": "G2032C"}, "param_type_id": "36301", "param_description": "A nucleotide or amino acid substitution that confers elevated resistance to antibiotic(s) relative to wild type. The most common type encoded in the CARD is an amino acid substitution gleaned from the literature with format [wild-type][position][mutation], e.g. R184Q. When present in the associated gene or protein, a single resistance variant confers resistance to an antibiotic drug or drug class. Single resistance variants are used by the protein variant and rRNA mutation models to detect antibiotic resistance from submitted sequences."}}, "ARO_description": "Point mutation in the 23S rRNA of Escherichia coli shown to confer resistance to oxazolidinone type antibiotics", "model_sequences": {"sequence": {"4164": {"dna_sequence": {"fmax": "240063", "fmin": "237159", "accession": "AE014075.1", "strand": "+", "sequence": "GGTTAAGCGACTAAGCGTACACGGTGGATGCCCTGGCAGTCAGAGGCGATGAAGGACGTGCTAATCTGCGATAAGCGTCGGTAAGGTGATATGAACCGTTATAACCGGCGATTTCCGAATGGGGAAACCCAGTGTGTTTCGACACACTATCATTAACTGAATCCATAGGTTAATGAGGCGAACCGGGGGAACTGAAACATCTAAGTACCCCGAGGAAAAGAAATCAACCGAGATTCCCCCAGTAGCGGCGAGCGAACGGGGAGGAGCCCAGAGCCTGAATCAGTGTGTGTGTTAGTGGAAGCGTCTGGAAAGGCGCGCGATACAGGGTGACAGCCCCGTACACAAAAATGCACATGCTGTGAGCTCGATGAGTAGGGCGGGACACGTGGTATCCTGTCTGAATATGGGGGGACCATCCTCCAAGGCTAAATACTCCTGACTGACCGATAGTGAACCAGTACCGTGAGGGAAAGGCGAAAAGAACCCCGGCGAGGGGAGTGAAAAAGAACCTGAAACCGTGTACGTACAAGCAGTGGGAGCATGCTTAGGCGTGTGACTGCGTACCTTTTGTATAATGGGTCAGCGACTTATATTCTGTAGCAAGGTTAACCGAATAGGGGAGCCGAAGGGAAACCGAGTCTTAACTGGGCGTTAAGTTGCAGGGTATAGACCCGAAACCCGGTGATCTAGCCATGGGCAGGTTGAAGGTTGGGTAACACTAACTGGAGGACCGAACCGACTAATGTTGAAAAATTAGCGGATGACTTGTGGCTGGGGGTGAAAGGCCAATCAAACCGGGAGATAGCTGGTTCTCCCCGAAAGCTATTTAGGTAGCGCCTCGTGAACTCATCTCCGGGGGTAGAGCACTGTTTCGGCAAGGGGGTCATCCCGACTTACCAACCCGATGCAAACTGCGAATACCGGAGAATGTTATCACGGGAGACACACGGCGGGTGCTAACGTCCGTCGNGAAGAGGGAAACAACCCAGACCGCCAGCTAAGGTCCCAAAGTCATGGTTAAGTGGGAAACGATGTGGGAAGGCCCAGACAGCCAGGATGTTGGCTTAGAAGCAGCCATCATTTAAAGAAAGCGTAATAGCTCACTGGTCGAGTCGGCCTGCGCGGAAGATGTAACGGGGCTAAACCATGCACCGAAGCTGCGGCAGCGACGCTTATGCGTTGTTGGGTAGGGGAGCGTTCTGTAAGCCTGTGAAGGTGTACTGTGAGGTATGCTGGAGGTATCAGAAGTGCGAATGCTGACATAAGTAACGATAAAGCGGGTGAAAAGCCCGCTCGCCGGAAGACCAAGGGTTCCTGTCCAACGTTAATCGGGGCAGGGTGAGTCGACCCCTAAGGCGAGGCCGAAAGGCGTAGTCGATGGGAAACAGGTTAATATTCCTGTACTTGGTGTTACTGCGAAGGGGGGACGGAGAAGGCTATGTTGGCCGGGCGACGGTTGTCCCGGTTTAAGCGTGTAGGCTGGTTTTCCAGGCAAATCCGGAAAATCAAGGCTGAGGCGTGATGACGAGGCACTACGGTGCTGAAGCAACAAATGCCCTGCTTCCAGGAAAAGCCTCTAAGCATCAGGTAACATCAAATCGTACCCCAAACCGACACAGGTGGTCAGGTAGAGAATACCAAGGCGCTTGAGAGAACTCGGGTGAAGGAACTAGGCAAAATGGTGCCGTAACTTCGGGAGAAGGCACGCTGATATGTAGGTGAAGCGACTTGCTCGTGGAGCTGAAATCAGTCGAAGATACCAGCTGGCTGCAACTGTTTATTAAAAACACAGCACTGTGCAAACACGAAAGTGGACGTATACGGTGTGACGCCTGCCCGGTGCCGGAAGGTTAATTGATGGGGTTAGCGCAAGCGAAGCTCTTGATCGAAGCCCCGGTAAACGGCGGCCGTAACTATAACGGTCCTAAGGTAGCGAAATTCCTTGTCGGGTAAGTTCCGACCTGCACGAATGGCGTAATGATGGCCAGGCTGTCTCCACCCGAGACTCAGTGAAATTGAACTCGCTGTGAAGATGCAGTGTACCCGCGGCAAGACGGAAAGACCCCGTGAACCTTTACTATAGCTTGACACTGAACATTGAGCCTTGATGTGTAGGATAGGTGGGAGGCTTTGAAGTGTGGACGCCAGTCTGCATGGAGCCGACCTTGAAATACCACCCTTTAATGTTTGATGTTCTAACGTTGACCCGTAATCCGGGTTGCGGACAGTGTCTGGTGGGTAGTTTGACTGGGGCGGTCTCCTCCTAAAGAGTAACGGAGGAGCACGAAGGTTGGCTAATCCTGGTCGGACATCAGGAGGTTAGTGCAATGGCATAAGCCAGCTTGACTGCGAGCGTGACGGCGCGAGCAGGTGCGAAAGCAGGTCATAGTGATCCGGTGGTTCTGAATGGAAGGGCCATCGCTCAACGGATAAAAGGTACTCCGGGGATAACAGGCTGATACCGCCCAAGAGTTCATATCGACGGCGGTGTTTGGCACCTCGATGTCGGCTCATCACATCCTGGGGCTGAAGTAGGTCCCAAGGGTATGGCTGTTCGCCATTTAAAGTGGTACGCGAGCTGGGTTTAGAACGTCGTGAGACAGTTCGGTCCCTATCTGCCGTGGGCGCTGGAGAACTGAGGGGGGCTGCTCCTAGTACGAGAGGACCGGAGTGGACGCATCACTGGTGTTCGGGTTGTCATGCCAATGGCACTGCCCGGTAGCTAAATGCGGAAGAGATAAGTGCTGAAAGCATCTAAGCACGAAACTTGCCCCGAGATGAGTTCTCCCTGACTCCTTGAGGGTCCTGAAGGAACGTTGAAGACGACGACGTTGATAGGCCGGGTGTGTAAGCGCAGCGATGCGTTGAGCTAACCGGTACTAATGAACCGTGAGGCTTAACCTT"}, "NCBI_taxonomy": {"NCBI_taxonomy_name": "Escherichia coli CFT073", "NCBI_taxonomy_id": "199310", "NCBI_taxonomy_cvterm_id": "36763"}, "protein_sequence": {"accession": "", "sequence": ""}}}}, "ARO_category": [], "ARO_name": "Escherichia coli 23S rRNA with mutation conferring resistance to oxazolidinone antibiotics", "model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "ARO_id": "41324", "model_name": "Escherichia coli 23S rRNA with mutation conferring resistance to oxazolidinone antibiotics", "model_type_id": "40295"}, "2818": {"model_id": "2818", "ARO_accession": "3004171", "model_param": {"blastn_bit_score": {"param_value": "5000", "param_type_id": "41093", "param_type": "BLASTN bit-score", "param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}, "snp": {"param_type": "single resistance variant", "param_value": {"7931": "A2305G"}, "clinical": {"7931": "A2305G"}, "param_type_id": "36301", "param_description": "A nucleotide or amino acid substitution that confers elevated resistance to antibiotic(s) relative to wild type. The most common type encoded in the CARD is an amino acid substitution gleaned from the literature with format [wild-type][position][mutation], e.g. R184Q. When present in the associated gene or protein, a single resistance variant confers resistance to an antibiotic drug or drug class. Single resistance variants are used by the protein variant and rRNA mutation models to detect antibiotic resistance from submitted sequences."}}, "ARO_description": "Point mutation in the 23S rRNA of Streptomyces ambofaciens shown to confer resistance to macrolide type antibiotics", "model_sequences": {"sequence": {"4163": {"dna_sequence": {"fmax": "4338141", "fmin": "4334992", "accession": "NZ_CP012949.1", "strand": "+", "sequence": "ACAGTGGACGCGAGCATCTGTGGCCAAGTTTTTAAGGGCGCACGGTGGATGCCTTGGCACCAGGAACCGATGAAGGACGTGGGAGGCCACGATAGTCCCCGGGGAGCCGTCAACCAGGCTTTGATCCGGGGGTTTCCGAATGGGGAAACCCGGCAGTCGTCATGGGCTGTCACCCGCTGCTGAACACATAGGCAGTGTGGAGGGAACGCGGGGAAGTGAAACATCTCAGTACCCGCAGGAAGAGAAAACAACCGTGATTCCGGGAGTAGTGGCGAGCGAAACCGGATGAGGCCAAACCGTATACGTGTGAGACCCGGCAGGGGTTGCGTGTGCGGGGTTGTGGGATCTCTCTTTCACAGTCTGCCGGCTGTGAGGCGAGTCAGAAACCGTTGATGTAGGCGAAGGACATGCGAAAGGTCCGGCGTAGAGGGTAAGACCCCCGTAGTCGAAACATCAGCGGCTCGTTTGAGAGACACCCAAGTAGCACGGGGCCCGAGAAATCCCGTGTGAATCTGGCGGGACCACCCGCTAAGCCTAAATATTCCCTGGTGACCGATAGCGGATAGTACCGTGAGGGAATGGTGAAAAGTACCGCGGGAGCGGAGTGAAATAGTACCTGAAACCGTGTGCCTACAAGCCGTGGGAGCGTCGGACATCAAGCTTGCTTGGTGTCTCGTGACTGCGTGCCTTTTGAAGAATGAGCCTGCGAGTTTGCGGTGTGTTGCGAGGTTAACCCGAGTGGGGAAGCCGTAGCGAAAGCGAGTCCGAATAGGGCGTTTCAGTAGCACGCTCAAGACCCGAAGCGGAGTGATCTAGCCATGGGCAGGTTGAAGCGGAGGTAAGACTTCGTGGAGGACCGAACCCACCAGGGTTGAAAACCTGGGGGATGACCTGTGGTTAGGGGTGAAAGGCCAATCAAACTCCGTGATAGCTGGTTCTCCCCGAAATGCATTTAGGTGCAGCGTCGTGTGTTTCTTGCCGGAGGTAGAGCACTGGATAGGCGATGGGCCCTACCGGGTTACTGACCTTAGCCAAACTCCGAATGCCGGTAAGTGAGAGCGCGGCAGTGAGACTGTGGGGGATAAGCTCCATGGTCGAGAGGGAAACAGCCCAGAGCATCGACTAAGGCCCCTAAGCGTACGCTAAGTGGGAAAGGATGTGGAGTCGCACAGACAACCAGGAGGTTGGCTTAGAAGCAGCCACCCTTGAAAGAGTGCGTAATAGCTCACTGGTCTAGTGATTCCGCGCCGACAATGTAGCGGGGCTCAAGCGTACCGCCGAAGTCGTGTCATTCATACAATAGGGCCAACGCCTGTATGGATGGGTAGGGGAGCGTCGTGTGCCGGGTGAAGCAGCCGCGGAAGCGAGTTGTGGACGGTTCACGAGTGAGAATGCAGGCATGAGTAGCGATACAAACGTGAGAAACGTTTGCGCCGATTGACTAAGGGTTCCTGGGTCAAGCTGATCTGCCCAGGGTAAGTCGGGACCTAAGGCGAGGCCGACAGGCGTAGTCGATGGATAACCGGTTGATATTCCGGTACCCGCTGTGAAGCGTCAAACATCGAGCATCGTGATGCTAAGGCCGTGAAGCCGCCCTGATCTCTTCGGAGTTGAGGGGAGTGGTGGAGCCGCCGGACCAAGCGGTTAGTAGGTGAGTGATGGGGTGACGCAGGAAGGTAGTCCATCCCGGGCGGTGGTTGTCCCGGGGTAAGGGTGTAGGACGTCAGGTAGGTAAATCCGCCTGGCAATAGTCTGAGACCTGATGCCGAGCCGATTGTGGTGAAGTGGATGATCCTATGCTGTCGAGAAAAGCCTCTAGCGAGTTTCATGGCGGCCCGTACCCTAAACCGACTCAGGTGGTCAGGTAGAGAATACCGAGGCGTTCGGGTGAACTATGGTTAAGGAACTCGGCAAAATGCCCCCGTAACTTCGGGAGAAGGGGGGCCACACCTGGTGATGATCTTTACGGTCTGAGCTGGGGGTGGCCGCAGAGACCAGCGAGAAGCGACTGTTTACTAAAAACACAGGTCCGTGCGAAGCCGTAAGGCGATGTATACGGACTGACGCCTGCCCGGTGCTGGAACGTTAAGGGGACCGGTTAGCTCCATTTCGGTGGGGCGAAGCTGAGAACTTAAGCGCCAGTAAACGGCGGTGGTAACTATAACCATCCTAAGGTAGCGAAATTCCTTGTCGGGTAAGTTCCGACCTGCACGAATGGCGTAACGACTTCTCGACTGTCTCAACCATAGGCCCGGTGAAATTGCACTACGAGTAAAGATGCTCGTTTCGCGCAGCAGGACGGAAAGACCCCGGGACCTTTACTACAGTTTGATATTGGTGTTCGGTTCGGCTTGTGTAGGATAGCTGGGAGACTTTGAAGCTCGCACGCCAGTGTGGGTGGAGTCGTCGTTGAAATACCAGTCTGGTCGTGCTGGATGTCTAACCTGGGTCCGTGATCCGGATCAGGGACAGTGTCTGATGGGTAGTTTAACTGGGGCGGTTGCCTCCTAAAGAGTAACGGAGGCGCCCAAAGGTTCCCTCAGCCTGGTTGGCAATCAGGTGTTGAGTGTAAGTGCACAAGGGAGCTTGACTGTGAGACCGACGGGTCGAGCAGGGACGAAAGTCGGGACTAGTGATCCGGCGGTGGCTTGTGGAAGCGCCGTCGCTCAACGGATAAAAGGTACCCCGGGGATAACAGGCTGATCTTCCCCAAGAGTCCATATCGACGGGATGGTTTGGCACCTCGATGTCGGCTCGTCGCATCCTGGGGCTGGAGTCGGTCCCAAGGGTTGGGCTGTTCGCCCATTAAAGCGGTACGCGAGCTGGGTTTAGAACGTCGTGAGACAGTTCGGTCCCTATCCGCTGTGCGCGTAGGAGTCTTGAGAAGGGCTGTCCCTAGTACGAGAGGACCGGGACGGACGAACCTCTGGTGTGCCAGTTGTCCTGCCAAGGGCATGGCTGGTTGGCTACGTTCGGGAGGGATAACCGCTGAAAGCATCTAAGCGGGAAGCCTGCTTCGAGATGAGGACTCCCACCCCCTTGAGGGGTTAAGGCTCCCAGTAGACGACTGGGTTGATAGGCCGGATCTGGAAGCACCGCAAGGTGTGGAGGTGACCGGTACTAATAGGCCGAGGGCTTGTCCTCAGTTG"}, "NCBI_taxonomy": {"NCBI_taxonomy_name": "Streptomyces ambofaciens", "NCBI_taxonomy_id": "1889", "NCBI_taxonomy_cvterm_id": "36864"}, "protein_sequence": {"accession": "", "sequence": ""}}}}, "ARO_category": {"36454": {"category_aro_name": "determinant of macrolide resistance", "category_aro_cvterm_id": "36454", "category_aro_accession": "3000315", "category_aro_description": "Enzymes, other proteins or other gene products shown clinically to confer resistance to macrolide antibiotics."}, "35950": {"category_aro_name": "antibiotic resistant gene variant or mutant", "category_aro_cvterm_id": "35950", "category_aro_accession": "0000031", "category_aro_description": "Resistance to antibiotics is often conferred by single nucleotide polymorphisms (SNPs) and other mutations in target genes."}}, "ARO_name": "Streptomyces ambofaciens 23S rRNA with mutation conferring resistance to macrolide antibiotics", "model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "ARO_id": "41322", "model_name": "Streptomyces ambofaciens 23S rRNA with mutation conferring resistance to macrolides", "model_type_id": "40295"}, "2833": {"model_id": "2833", "ARO_accession": "3004133", "model_param": {"blastn_bit_score": {"param_value": "2000", "param_type_id": "41093", "param_type": "BLASTN bit-score", "param_description": "A score is a numerical value that describes the overall quality of an alignment. Higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. This parameter is used by AMR detection models without a protein reference sequence but including a nucleotide reference sequence, e.g. the rRNA gene variant model. The BLASTN bit-score parameter is a curated value determined from BLASTN analysis of the canonical nucleotide reference sequence of a specific AMR-associated gene against the database of CARD reference sequences. This value establishes a threshold for computational prediction of a specific gene amongst a batch of submitted sequences."}, "snp": {"param_type": "single resistance variant", "param_value": {"8198": "A2058G", "8199": "A2058U"}, "clinical": {"8198": "A2058G", "8199": "A2058U"}, "param_type_id": "36301", "param_description": "A nucleotide or amino acid substitution that confers elevated resistance to antibiotic(s) relative to wild type. The most common type encoded in the CARD is an amino acid substitution gleaned from the literature with format [wild-type][position][mutation], e.g. R184Q. When present in the associated gene or protein, a single resistance variant confers resistance to an antibiotic drug or drug class. Single resistance variants are used by the protein variant and rRNA mutation models to detect antibiotic resistance from submitted sequences."}}, "ARO_description": "Point mutation in B. hyodysenteriae 23S rRNA shown to confer resistance to tylosin and erythromycin", "model_sequences": {"sequence": {"4183": {"dna_sequence": {"fmax": "2515808", "fmin": "2512812", "accession": "NZ_CP015910.2", "strand": "-", "sequence": "ATATGGTCAAGTCATTGGTCTGATTAGTGCTGCTCAGCTGAACAGGTATTTCTTCCCTGCTTACACTTGCAGTCTATCAACGTCGTAGTCTCCAACGAAACTCATAGGGAAAGTTAATCTTGAAGGAGGCTTCCCACTTAGATGCTTTCAGCGGTTATCCCGTCCGCACATAGCTACTCTGCGATGCTCTTGGCAGAACAACAGATACACCAGCGGTGCGTTCATTCCGGTCCTCTCGTACTAAGAATGACTCTTCTCAACTTTCCAACGCCCACAACGGATAGGGACCAAACTGTCTCACGACGTTCTGAACCCAGCTCGCGTACCGCTTTAATTGGCGAACAGCCAAACCCTTGGGACCTGCTCCAGCCCCAGGATGCGATGAGCCGACATCGAGGTGCCAAACCTCCCCGTCGATATGAACTCTTGGGGGAGATAAGCCTGTTATCCCCGGAGTACCTTTTGTCCGTTTAGCGATGGCCCTTCCACTCGGGACCACCGGATCACTAAAACCTACTTTCGTACCTGCTCGAGATGTCTCTCTCGCAGTCAAGCCACCTTATGCCTTTATACTCTACTACCGATTTCTATTCGGTTTGAGGTGACCTTTGCACGCCTCCGTTACTCTTTAGGAGGCGACCGCCCCAGTCAAACTACCCGCCTGACAATGTCCGACTGCCGGATAACGGCTAATCGTTAGAATCCCATTTTGCGAAGGATGGTATTTCAACGATGGCTCCATGAAAGCTGGCGCTCCCACTTCAAAGCCTCCCATCTATCCTACACATCACAAAACAAAACTCAATGTCAAGTTATAGTAAAGGTTCACGGGGTCTTTCCGTCCTGTTGCGGGTAATCAGCATCTTCACTGATAATTCAATTTCACCGAGTTCTTCCCCGAGACAGTGCCCGGATCGTTACACCATTCGTGCAGGTCGGAACTTACCCGACAAGGAATTTCGCTACCTTAGGACCGTCATAGTTACGGCCGCCGTTTACTGGGGCTTCAATTCGAAGCTTCGCTTACGCTAACCTCTCCTTTTAACCTTCCAGCACTGGGCAGGTGTCAATCCCTATACATCCATTTACATGTTTGCAGAGATCTGTGTTTTTGTTAAACAGTCGGCCAGGCCTTTTCACTGCGACTCTCCTCTCAATATTGCTACCAAGATTCAAGCGTCTCTTTTTCCGAAGTTACAAGACTAATTTGCAGAGTTCCTTAGGGAAGATTATCTCGAGCGCCTTAGAATACTCATCTCACCCACCTGTGTCGGTTTGCGGTACGATTATGACATGCCTAACCTTAGAAATTATTTCTCGACAGCCTAGCTCATGCAACTTCCTGAACCCGAAAGCCCAGTCACTATCCAGCCTCAACCTTAAAGAAGCAAGCATTTGACTCACCTCAAGTCTAAACTGTTTGACGTTATCTACCAATCTAACGCGTACATAAACAATCTGTGTCATTCCATCGAAACATATCATAGTACAGGAATATTTACCTGTTTCCCATCGACTACGCTTTTCAGCCTCATCTTAGGGGTCGACTAACCCTAGGCAGATTAGCTTTACCTAGGAAACCTTGGGTTTGCGGCGAACGGGTTTCTCACCCGTTTTCTCGTTACTCATGCCTGCATCCTCACTTCTTATACCTCCAGCCCACCTCACGATGAACCTTCGACGGCTTAAAGAACGCTCTCCTACCAATTATAGATTAACTCTATAATTCCCTAGCTTCGGTACTATGTTTGAGCCCCGTTACATTTTCGGCGCAAGAACACTCGACCAGTGAGCTGTTACGCACTCTTTAAAGGAATGGCTGCTTCTAAGCCAACCTCCTGGCTGTTTAAGTATTCTCACATCCTTTCCCACTTAACATAGATTTTGGGACCTTAGCTGAGGATCTGGGCTGTTTCCCTTTTGACAATGACGCTTATCCGCCGCTGTCTAACTGCCATGTTCTTAACTTACGGTATTCGGAGTTTAGTTGGGTTTGGTACTCGGTTAGGAGCCCTAGTCCATTTAGTGCTCTACCCCCGCAAGTAAACACATAACGCTGCCCCTAAAGACATTTCGGAGAGAACCAGCTATCTCCAAGTTTGATTAGCCTTTCACTCCTACCCACAAGTCATCCAAAGCCTTTTCACGGCCACTGGTTCGCACCTCCATTCGATGTTACTCGAGTTTCGCGCTGCTCATAGGTAGATCACTTGGCTTCGGGTCGTATAGCATGCAACTAATTTCGCCCTATTAAGGCTCGCTTTCACTACGACTACAAGGCTATTACCTCTTAATCTTGCTACATACTATAAGTCGCAGGCTCATTCTACAAAAGGCACGCCATCAGGCCATCTACTATTGCTAGCAGGGTACCCTCTGACTACTTGTAAGCTTAAGGTTTCAGGTTCTATTTAATAACCCTCAACGGGAGACTTTTCACCTTTCCCTCACGGTACTCTACACTATTGGTCACTGGTTAGTATTTTGCCTTGGATAGTGGTCTACCCAGATTCAAACAGGGTTTCACGTGCCCCGCCCTACTCAGGAACGATAAAACCTTGTCTATATGATTTCGTGTACAGGACTATCACCCACTATGGTCTGCTTTTCCAAAACAGTTCCACTATCATATAAAGCAAAGTCGATGCATGCAAGTACATCGCTTACCGCCCTACAACACCATAATAACAACGCCTTGCAGCTTGACATTATTAAGGTTTAGGCTCTTCCCCTTTCGCTCGCCACTACTTAGGGAATCTCAATTTTGATTACTCTTCCTCCAGGTACTTAGATGTTTCAGTTCCCTGGGTGTCGCCTCATACACCTATGGATTCAGTGTATGATATAGAAGGTTTACTTCTATAGGTTTTCCCATTCGGTGATCTACGGATCATAGAATATTTGCTTCTCCCCGTAGCTTATCGCAGCTTGTCACGACCTTCATCGCCTTCCAGTGCCTAGGCATCCACCTTAAGCCCTTACTTACTTGACCATAT"}, "NCBI_taxonomy": {"NCBI_taxonomy_name": "Brachyspira hyodysenteriae ATCC 27164", "NCBI_taxonomy_id": "1266923", "NCBI_taxonomy_cvterm_id": "41351"}, "protein_sequence": {"accession": "", "sequence": ""}}}}, "ARO_category": {"36454": {"category_aro_name": "determinant of macrolide resistance", "category_aro_cvterm_id": "36454", "category_aro_accession": "3000315", "category_aro_description": "Enzymes, other proteins or other gene products shown clinically to confer resistance to macrolide antibiotics."}, "35950": {"category_aro_name": "antibiotic resistant gene variant or mutant", "category_aro_cvterm_id": "35950", "category_aro_accession": "0000031", "category_aro_description": "Resistance to antibiotics is often conferred by single nucleotide polymorphisms (SNPs) and other mutations in target genes."}}, "ARO_name": "Brachyspira hyodysenteriae 23S rRNA with mutation conferring resistance to tylosin", "model_type": "rRNA gene variant model", "model_description": "The rRNA gene variant model is an AMR detection model used to identify ribosomal RNA (rRNA) genes with mutations shown clinically to confer resistance to known antibiotic(s) relative to the wild-type rRNA sequence. Like the protein variant model, rRNA gene variant models detect the presence of an rRNA sequence based on its homolog, and then secondarily search submitted query sequences for a curated mutation. This model includes an rRNA gene reference sequence, a BLASTN bitscore cutoff, and a set of mapped resistance variants. A submitted sequence must have both high homolog to the reference sequence and include a known resistance variant to be detected.", "ARO_id": "41260", "model_name": "Brachyspira hyodysenteriae 23S rRNA with mutation conferring resistance to tylosin", "model_type_id": "40295"}, "2832": {"model_id": "2832", "ARO_accession": "3002527", "model_param": {"blastp_bit_score": {"param_value": "350", "param_type_id": "40725", "param_type": "BLASTP bit-score", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}, "ARO_description": "AAC(2')-Ie is a chromosomal-encoded aminoglycoside acetyltransferase in M. leprae", "model_sequences": {"sequence": {"4182": {"dna_sequence": {"fmax": "3039740", "fmin": "3039191", "accession": "NC_002677.1", "strand": "+", "sequence": "ATGGATACCCACCACGTACATACCGCCCGCCTGGTTCATACCGCCGATCTCGACGGCGAGACTCTCCGGCGCCTCCAGCAGATGGTCACCGACGCTTTCGCCGGAGACTTCGACGAGACCGACTGGGAACACGCATTGGGCGGGATGCACGCCCTGATCTGGCGGCACGGCACAATCATTGCGCACGCCGCGGTCGTTCAGCGGCGACTATTCTACCACGGCAACGCGTTGCGTTGCGGTTACCTCGAAGGTGTCGCGGTACGGAAGGACTGCCGGGGCCGCGGACTTGTACATGCGCTGCTGGACGCTATCGAGCAAGTGATACGCGGCGCCTATCAATTCGGTGCCTTGAGTTCCTCGGACCGAGCCCGTCGGGTGTATATGTCGCGCGGGTGGTTACCATGGCTCGGCCCGACGTCAGTGCTGGCTCCTACCGGTGTGATCCGTACGCCCGATGATGACGGCTCGGTGTTTGTCCTTCCGGTCGGCATCAACCCGGACACCTCCTCGGGGTTGATGTGCGATTGGCGCGCAGGCAACGTGTGGTAA"}, "NCBI_taxonomy": {"NCBI_taxonomy_name": "Mycobacterium leprae TN", "NCBI_taxonomy_id": "272631", "NCBI_taxonomy_cvterm_id": "40083"}, "protein_sequence": {"accession": "NP_302635.1", "sequence": "MDTHHVHTARLVHTADLDGETLRRLQQMVTDAFAGDFDETDWEHALGGMHALIWRHGTIIAHAAVVQRRLFYHGNALRCGYLEGVAVRKDCRGRGLVHALLDAIEQVIRGAYQFGALSSSDRARRVYMSRGWLPWLGPTSVLAPTGVIRTPDDDGSVFVLPVGINPDTSSGLMCDWRAGNVW"}}}}, "ARO_category": {"36696": {"category_aro_name": "antibiotic inactivation enzyme", "category_aro_cvterm_id": "36696", "category_aro_accession": "3000557", "category_aro_description": "Enzyme that catalyzes the inactivation of an antibiotic resulting in resistance.  Inactivation includes chemical modification, destruction, etc."}, "36243": {"category_aro_name": "determinant of aminoglycoside resistance", "category_aro_cvterm_id": "36243", "category_aro_accession": "3000104", "category_aro_description": "Enzymes, other proteins or other gene products shown clinically to confer resistance to aminoglycoside antibiotics."}}, "ARO_name": "AAC(2')-Ie", "model_type": "protein homolog model", "model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "ARO_id": "38927", "model_name": "AAC(2')-Ie", "model_type_id": "40292"}, "2831": {"model_id": "2831", "ARO_accession": "3000501", "model_param": {"blastp_bit_score": {"param_value": "2200", "param_type_id": "40725", "param_type": "BLASTP bit-score", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}, "ARO_description": "Due to gene duplication, the genomes of Nocardia species include both rifampin-sensitive beta-subunit of RNA polymerase (rpoB) and rifampin-resistant beta-subunit of RNA polymerase (rpoB2) genes, with ~88% similarity between the two gene products. Expression of the rpoB2 variant results in replacement of rifampin sensitivity with rifampin resistance.", "model_sequences": {"sequence": {"4176": {"dna_sequence": {"fmax": "4838688", "fmin": "4835199", "accession": "AP006618.1", "strand": "+", "sequence": "GTGCTGGAAGGACGCATCGTGACAGTCTCGAGTCGGACCGAGTCCCCGCTCGCCGCGCCCGGAGTCCCGGGTGCGCCCCGGCGGCTCTCGTTCGCCAGAATCCGCGAGCCCCTGGCGGTTCCGGGCCTGCTGGACATTCAGACCGAATCGTTCGGGTGGCTGATCGGCGCACCCGACTGGTGCGCCCGCGCGGCAGCCCGCGGGACGGAACCGGTCGCCGGGCTGGCCGAGGTGCTCGCCGAGATCAGCCCCATCGAGGACTTCGCGGGCACGATGTCGCTGACGTTGTCCGATCCGCGGTTCGAGGAGGTCAAGGCCTCCGTCGAGGAGTGCAAGGACAAGGATCTGACCTATGCCGCGCCGTGGTTCGTCACGGCGGAATTCGTGAACAACAACACCGGCGAGATCAAGAGCCAGACCGTTTTCATGGGTGATTTCCCCATGATGACCGCGCACGGGACATTCGTCGTCAACGGCACCGAACGGGTGGTCGTGTCGCAACTGGTGCGCTCGCCGGGCGTGTACTTCGACCACGCGATCGACAAGGGTTCGGAGAAGGACGTGCACAGCGCGCGGGTGATCCCGTCGCGCGGTGCCTGGCTGGAGTTCGACGTCGACAAGCGCGACACCCTGGGCGTGCGCATCGACCGCAAGCGCCGTCAGCCGGTAACGGTGCTGCTCAAGGCGCTGGGCTGGAGTGCGGAGCGAATCGCCGAGCGATTCGGGTTCGCCCCGCTGATCATGGCCTCGCTGGCCAAGGACAACGTGGCGGGCACCGACGACGCTTTGCTCGAGATCCATCGCAAGCTGCGTCCGGGCGAGCCGCCGACCAAGGAGTCGGCACAGAACCTGCTGGCAAATCTGTTCTTCACCGAGAAGCGCTACGACCTGGCGCGGGTGGGCCGTTACAAGATCGACAAGAAGCTCGGGCTGCGCGCTCCGGGGGCACCGCGGGTGCTCACCGAGGACGACATCGCGGCCACCATCGAGTACCTGGTGCGCCTGCACGCCGGGGAGAGGACGATGATCGCGCCCGGCGGGGTCGAGGTGCCGGTGGAGGTCGACGACATCGACCATTTCGGTAACCGCCGGGTGCGCACCGTGGGTGAGCTGATCCAGAACCAGATCCGCGTGGGTCTGTCCCGGATGGAACGGGTGGTGCGGGAGCGCATGACCACCCAGGACGTCGAGGCGATCACGCCGCAGTCGCTGATGAACATCCGCCCGGTGGTCGCGGCGATGAAGGAGTTCTTCGGCACCTCGCAGCTGTCGCAGTTCATGGACCAGCGCAACCCGCTGGCGAGCCTGACCAACAAGCGGCGGCTGTCCGCGCTCGGGCCGGGCGGGCTCTCGCGCGAACGGGCCGGCTTGGAGGTCCGCGACGTGCACTACAGCCACTACGGCCGGATGTGCCCGATCGAGACGCCGGAGGGGCCGAACATCGGCCTGATGGGGTACCTGTCGGTGTACGCGCGGGTCAACCCGTTCGGTTTCGTCGAGACCCCCTACCGGCGGGTGGTGGACGGCCGGGTCACCGACGAGGTCGATTACCTCACCGCCGACGAGGAGGACCGGCACGTGGTGGCCCAGGCGAACGAACCGCTGGACGCCGAGGGCCGCTTCCTGGCCGCGCGAATCCCGGTGCGCCGGAAGAACTCCGAGGTGGAACTCGTCGATTCCGCCGCGGTGGATTACATGGATGTGTCGCCGCGGCAGATGGTGTCGGTGGCGACGGCGATGATTCCGTTCCTCGAGCATGATGATGCCAACCGTGCGCTGATGGGTGCGAATATGCAGCGTCAGGCGGTGCCGTTGATTCGGTCGGAGGCGCCGATCGTGGGTACCGGTATGGAGCTGCGTGCGGCGGTGGATGCCGGTGATGTGGTGGTGAACGAGAAGGCCGGTGTGGTCGAGGAGGTTTCGGCCGATTACGTCACGGTGATGGCCGATGACGGGACTCGTAAGTCGTATCGGATGCGGAAGTTCAACCGGTCGAATCAGGGGACGTGTTCGAATCAGCGGCCGATCGTGGACGAGGGTCAGCGGGTCGAGGCCGGGCAGGTGTTGGCTGATGGGCCGTGCACCGAGAACGGTGAGATGGCGCTGGGTAAGAACCTGTTGGTGGCGATCATGCCGTGGGAGGGGCACAACTACGAGGACGCGATCATCCTGTCGCAGCGGTTGGTGGAGCAGGATGTGTTGACCTCGATCCATATCGAGGAGCACGAGATCGATGCTCGTGACACCAAGCTCGGTGCCGAGGAGATCACTCGCGACATTCCGAATGTCTCCGATGAGGTGTTGGCGGATCTGGACGAGCGTGGCATCGTGCGGATCGGTGCGGAGGTGCGTGATGGTGACATCCTGGTCGGTAAGGTCACGCCCAAGGGTGAGACCGAGCTGACGCCGGAGGAGCGGTTGCTGCGGGCGATCTTCGGGGAGAAGGCGCGTGAGGTGCGTGATACCTCGTTGAAGGTGCCCCACGGTGAGTCGGGCAAGGTCATCGGGATCCGGGTGTTCTCTCGGGAGGACGATGACGATCTGCCGCCCGGTGTCAACGAGTTGGTGCGGGTGTATGTGGCGCAGAAGCGCAAGATCCAAGACGGTGACAAGCTGGCCGGTCGGCACGGGAACAAGGGTGTGATCGGCAAGATCCTGCCCACCGAGGACATGCCGTTCCTGCCCGATGGCACCCCGGTCGACATCATCCTCAACACCCACGGTGTGCCGCGGCGTATGAACATCGGCCAGATCCTGGAAACCCACCTCGGCTGGATCGGCAAAGCCGGCTGGAAGGTCGAGGGCAACCCCGAGTGGGCCAAGGATCTGCCGGAGGAGATGTGGGAGGCCCCCGCGGACTCCAACATCGCGACTCCGGTGTTCGACGGTGCCCGCGAGGAGGAGCTGACGGGTCTGCTGGGCTCGACGCTGCCCAACCGTGACGGTGAGCGGATGGTCGACGACAACGGCAAGGCCGTGCTGTTCGACGGTCGTTCCGGTGAGCCGTTCCCGTACCCGGTGGCGGTGGGCTACATGTACATCCTGAAGCTGCACCACCTGGTCGACGACAAGATCCACGCTCGCTCCACCGGCCCGTACTCGATGATCACCCAGCAGCCGCTCGGTGGTAAGGCCCAGTTCGGTGGCCAGCGCTTCGGTGAGATGGAGTGCTGGGCGATGCAGGCCTACGGCGCGGCCTACACCCTGCAGGAACTGCTCACGATCAAGTCCGACGACGTCGTCGGCCGCGTGAAGGTCTACGAGGCGATCGTCAAGGGCGACAACATCCCGGAACCCGGTGTGCCGGAATCGTTCAAGGTGTTGTTGAAGGAACTCCAGGCACTGTGCCTCAACGTCGAGGTGCTGTCCGCCGGTGCCGCGGTCGAACTGGCGCACGGGGTCGACGACGATCACGAGCGGACGGCCGCGAACCTCGGCATCAACCTGTCCCGCGCCGAATCCATCACGGAGACCGAGCTGTCCGGGTGA"}, "NCBI_taxonomy": {"NCBI_taxonomy_name": "Nocardia farcinica IFM 10152", "NCBI_taxonomy_id": "247156", "NCBI_taxonomy_cvterm_id": "41346"}, "protein_sequence": {"accession": "BAD59497.1", "sequence": "MLEGRIVTVSSRTESPLAAPGVPGAPRRLSFARIREPLAVPGLLDIQTESFGWLIGAPDWCARAAARGTEPVAGLAEVLAEISPIEDFAGTMSLTLSDPRFEEVKASVEECKDKDLTYAAPWFVTAEFVNNNTGEIKSQTVFMGDFPMMTAHGTFVVNGTERVVVSQLVRSPGVYFDHAIDKGSEKDVHSARVIPSRGAWLEFDVDKRDTLGVRIDRKRRQPVTVLLKALGWSAERIAERFGFAPLIMASLAKDNVAGTDDALLEIHRKLRPGEPPTKESAQNLLANLFFTEKRYDLARVGRYKIDKKLGLRAPGAPRVLTEDDIAATIEYLVRLHAGERTMIAPGGVEVPVEVDDIDHFGNRRVRTVGELIQNQIRVGLSRMERVVRERMTTQDVEAITPQSLMNIRPVVAAMKEFFGTSQLSQFMDQRNPLASLTNKRRLSALGPGGLSRERAGLEVRDVHYSHYGRMCPIETPEGPNIGLMGYLSVYARVNPFGFVETPYRRVVDGRVTDEVDYLTADEEDRHVVAQANEPLDAEGRFLAARIPVRRKNSEVELVDSAAVDYMDVSPRQMVSVATAMIPFLEHDDANRALMGANMQRQAVPLIRSEAPIVGTGMELRAAVDAGDVVVNEKAGVVEEVSADYVTVMADDGTRKSYRMRKFNRSNQGTCSNQRPIVDEGQRVEAGQVLADGPCTENGEMALGKNLLVAIMPWEGHNYEDAIILSQRLVEQDVLTSIHIEEHEIDARDTKLGAEEITRDIPNVSDEVLADLDERGIVRIGAEVRDGDILVGKVTPKGETELTPEERLLRAIFGEKAREVRDTSLKVPHGESGKVIGIRVFSREDDDDLPPGVNELVRVYVAQKRKIQDGDKLAGRHGNKGVIGKILPTEDMPFLPDGTPVDIILNTHGVPRRMNIGQILETHLGWIGKAGWKVEGNPEWAKDLPEEMWEAPADSNIATPVFDGAREEELTGLLGSTLPNRDGERMVDDNGKAVLFDGRSGEPFPYPVAVGYMYILKLHHLVDDKIHARSTGPYSMITQQPLGGKAQFGGQRFGEMECWAMQAYGAAYTLQELLTIKSDDVVGRVKVYEAIVKGDNIPEPGVPESFKVLLKELQALCLNVEVLSAGAAVELAHGVDDDHERTAANLGINLSRAESITETELSG"}}}}, "ARO_category": {"36522": {"category_aro_name": "determinant of rifamycin resistance", "category_aro_cvterm_id": "36522", "category_aro_accession": "3000383", "category_aro_description": "Enzymes, other proteins, or other gene products shown clinically to confer resistance to rifamycin (rifampin) antibiotics."}, "35950": {"category_aro_name": "antibiotic resistant gene variant or mutant", "category_aro_cvterm_id": "35950", "category_aro_accession": "0000031", "category_aro_description": "Resistance to antibiotics is often conferred by single nucleotide polymorphisms (SNPs) and other mutations in target genes."}, "36520": {"category_aro_name": "antibiotic target replacement protein", "category_aro_cvterm_id": "36520", "category_aro_accession": "3000381", "category_aro_description": "Alternate proteins that have the same functions as other antibiotic target proteins, but are structurally different and thus resistant to antibiotics. These can replace the activity of other antibiotic-sensitive proteins in the presence of antibiotics."}}, "ARO_name": "Nocardia rifampin resistant beta-subunit of RNA polymerase (rpoB2)", "model_type": "protein homolog model", "model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "ARO_id": "36640", "model_name": "Nocardia rifampin resistant beta-subunit of RNA polymerase (rpoB2)", "model_type_id": "40292"}, "2830": {"model_id": "2830", "ARO_accession": "3000248", "model_param": {"blastp_bit_score": {"param_value": "500", "param_type_id": "40725", "param_type": "BLASTP bit-score", "param_description": "A score is a numerical value that describes the overall quality of an alignment with higher numbers correspond to higher similarity. The bit-score (S) is determined by the following formula: S = (\u03bb \u00d7 S \u2212 lnK)/ ln2 where \u03bb is the Gumble distribution constant, S is the raw alignment score, and K is a constant associated with the scoring matrix. Many AMR detection models use this parameter, including the protein homolog and protein variant models. The BLASTP bit-score parameter is a curated value determined from BLASTP analysis of the canonical reference sequence of a specific AMR-associated protein against the database of CARD reference sequence. This value establishes a threshold for computational prediction of a specific protein amongst a batch of submitted sequences."}}, "ARO_description": "DnaA is a chromosomal replication initiation protein which binds and interacts with RNA polymerase in Escherichia coli. A surplus of DnaA present in a cell has been shown to confer resistance to the antibiotic Rifampicin. Normally, rifampicin inhibits initiation of transcription by RNA polymerase, but a surplus of DnaA available at the origin has been shown to disrupt Rifampicin activity and confer resistance.", "model_sequences": {"sequence": {"4175": {"dna_sequence": {"fmax": "3883729", "fmin": "3882325", "accession": "NC_000913.3", "strand": "-", "sequence": "TTACGATGACAATGTTCTGATTAAATTTGAAAAATCTTCTTTGATATCGTGGCTCTCTTCACGCAACTGCTCGATCTTACGGCAGGCATGAAGCACCGTCGTGTGGTCACGGCCACCAAACGCATCGCCAATCTCCGGCAGACTGTGGTTAGTCAGCTCTTTCGCCAGCGCCATCGCCATCTGGCGCGGACGCGCCACCGAGCGGGATCGACGCTTGGAAAGGAGATCCGCGACTTTGATCTTGTAGTACTCCGCCACCGTCTTCTGAATATTGTCGATGGTGACCAGTTTTTCCTGCAATGCCAGCAAGTCGCGCAGCGCCTCACGCACGAAGTCGATGGTGATCGCCCGTCCGGTAAAGTTGGCATTGGCAATGACGCGGTTCAGCGCCCCTTCCAGCTCACGTACGTTAGATCGTAGACGCTTGGCGATAAAGAACGCCACTTCGCCCGGCAAACGAATGTCGTTTTCGTCGGCCTTTTTCATCAGGATCGCCACACGGGTTTCCAGCTCTGGCGGTTCGATCGCCACAGTCAGTCCCCAACCGAAGCGGGATTTCAAACGATCCTCAACGCCGTTGATCTCTTTCGGATAGCGATCCGAGGTGAGAATGATCTGTTGATTACCTTCCAGCAGGGCGTTGAAGGTGTGGAAAAACTCTTCCTGAGATCGTTCTTTATTAGCAAAAAACTGAATATCGTCGATCAGCAGTGCATCTACGGAACGGTAGTAGCGTTTAAACTCTTCGATCGCGTTGTTTTGCAGGGCTTTAACCATGTCCTGAACAAAGCGCTCGGAGTGCATATAAACCACTTTGGCATTCGGCTTGCGCGCCATAATGCCGTTACCCACCGCATGCAGCAGGTGAGTTTTACCCAGACCCGTGCCGCCATAAAGGAACAACGGGTTATAGGCACCGCCAGGGTTATCCGCCACCTGGCGAGCCGCCGCGCGCGCCAGTTGGTTAGATTTACCTTCAACGAAGTTATCAAACGTGTGTTTGACGTTTACGTTAGAACGATAGGTCGGTTCTGCCGGGGCCGGGACGTTATCCCAACCTGAGCGCGTAGAAGGCGCAGCACGTTGCGGCTGCGTTTGCGCCACCTGTGCAGGGGCCGCGACGTTGCTCGTCACTGCCGCTTGTGGCGTTTGCGTCACCGGTTTGGTGCCGACTTCAAAACGCAGCTGTGGGGCATCCGCTCCGCAGAAACTGGTTAGCAGTCCATTGATATTATTAAGGTACTTGTCCCGTACCCAATCGAGGACAAAACGGTTTGGCGCGTACAGGGCCAGCGTGTTATCGCTCAGTTCCGCCTGCAATGGGCGTATCCACATACTGAATTCTGTGGCTGGTAACTCATCCTGCAATCGGGCAAGACACTGCTGCCAAAGCGAAAGTGACAC"}, "NCBI_taxonomy": {"NCBI_taxonomy_name": "Escherichia coli str. K-12 substr. MG1655", "NCBI_taxonomy_id": "511145", "NCBI_taxonomy_cvterm_id": "36849"}, "protein_sequence": {"accession": "NP_418157.1", "sequence": "MSLSLWQQCLARLQDELPATEFSMWIRPLQAELSDNTLALYAPNRFVLDWVRDKYLNNINGLLTSFCGADAPQLRFEVGTKPVTQTPQAAVTSNVAAPAQVAQTQPQRAAPSTRSGWDNVPAPAEPTYRSNVNVKHTFDNFVEGKSNQLARAAARQVADNPGGAYNPLFLYGGTGLGKTHLLHAVGNGIMARKPNAKVVYMHSERFVQDMVKALQNNAIEEFKRYYRSVDALLIDDIQFFANKERSQEEFFHTFNALLEGNQQIILTSDRYPKEINGVEDRLKSRFGWGLTVAIEPPELETRVAILMKKADENDIRLPGEVAFFIAKRLRSNVRELEGALNRVIANANFTGRAITIDFVREALRDLLALQEKLVTIDNIQKTVAEYYKIKVADLLSKRRSRSVARPRQMAMALAKELTNHSLPEIGDAFGGRDHTTVLHACRKIEQLREESHDIKEDFSNLIRTLSS"}}}}, "ARO_category": {"36324": {"category_aro_name": "antibiotic target protection protein", "category_aro_cvterm_id": "36324", "category_aro_accession": "3000185", "category_aro_description": "These proteins confer antibiotic resistance by bind the antibiotic target to prevent antibiotic binding."}, "36522": {"category_aro_name": "determinant of rifamycin resistance", "category_aro_cvterm_id": "36522", "category_aro_accession": "3000383", "category_aro_description": "Enzymes, other proteins, or other gene products shown clinically to confer resistance to rifamycin (rifampin) antibiotics."}}, "ARO_name": "DnaA", "model_type": "protein homolog model", "model_description": "The protein homolog model is an AMR detection model. Protein homolog models detect a protein sequence based on its similarity to a curated reference sequence. A protein homolog model has only one parameter: a curated BLASTP bitscore cutoff for determining the strength of a match. Protein homolog model matches to reference sequences are categorized on three criteria: \"perfect\", \"strict\" and \"loose\". A perfect match is 100% identical to the reference sequence along its entire length; a strict match is not identical but the bitscore of the matched sequence is greater than the curated BLASTP bitscore cutoff. Loose matches are other sequences with a match bitscore less than the curated BLASTP bitscore.", "ARO_id": "36387", "model_name": "DnaA", "model_type_id": "40292"}}}