{"id":"W3215832582","doi":"10.1099/mgen.0.000728","title":"ECTyper: in silico Escherichia coli serotype and species prediction from raw and assembled whole-genome sequence data","year":2021,"lang":"en","type":"article","venue":"Microbial Genomics","topic":"Escherichia coli research studies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":131,"is_retracted":false,"has_abstract":true,"ca_institutions":"Canadian Food Inspection Agency; Public Health Agency of Canada","funders":"Public Health Agency of Canada","keywords":"In silico; Serotype; Escherichia coli; Whole genome sequencing; Computational biology; Genome; Sequence (biology); Biology; Genetics; Raw data; Microbiology; Gene; Mathematics; Statistics","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.003371823,0.002632439,0.001150499,0.001782452,0.0005242674,0.00158501,0.001360631,0.0009355057,0.00339091],"category_scores_gemma":[0.006593344,0.0007955601,0.002543136,0.001063898,0.0003094035,0.000895692,0.00133311,0.001328547,0.00314565],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0004426315,"about_ca_system_score_gemma":0.001306583,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003019425,"about_ca_topic_score_gemma":0.005247462,"domain_scores_codex":[0.9983631,0.0004732275,0.0001799223,0.0005479712,0.0002994619,0.000136384],"domain_scores_gemma":[0.998225,0.001000769,0.0001562412,0.0002871705,0.0002425637,0.00008832705],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00585791,0.001277549,0.1946836,0.00578902,0.003799654,0.002735483,0.001600426,0.2517727,0.2028163,0.005023296,0.1197485,0.2048956],"study_design_scores_gemma":[0.0003715055,0.0008424328,0.03837329,0.0003041984,0.0004470476,0.001023906,0.0004881643,0.8276818,0.08285506,0.004808929,0.04258465,0.0002190355],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"methods","genre_scores_codex":[0.529359,0.001574978,0.2578201,0.0005688648,0.0005291073,0.0006717468,0.111547,0.09127221,0.006656982],"genre_scores_gemma":[0.3979521,0.0005754066,0.3859517,0.0005128276,0.00008125177,0.0005740407,0.2066026,0.005563505,0.002186484],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.00339091,"threshold_uncertainty_score":0.0178321,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03759364213199561,"score_gpt":0.2697008992175524,"score_spread":0.2321072570855567,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}