{"id":"W2811501037","doi":"10.1016/j.meegid.2018.06.029","title":"Towards better prediction of Mycobacterium tuberculosis lineages from MIRU-VNTR data","year":2018,"lang":"en","type":"article","venue":"Infection Genetics and Evolution","topic":"Tuberculosis Research and Epidemiology","field":"Medicine","cited_by":11,"is_retracted":false,"has_abstract":true,"ca_institutions":"Simon Fraser University","funders":"National Institute of General Medical Sciences; Natural Sciences and Engineering Research Council of Canada","keywords":"Computer science; Genotyping; Set (abstract data type); Artificial intelligence; Machine learning; Typing; Mycobacterium tuberculosis; Training set; Data mining; Tuberculosis; Biology; Genotype; Genetics","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0003296,0.0001164531,0.0002332531,0.0001394547,0.00009015469,0.00001368423,0.000061707,0.0001847737,0.0001504026],"category_scores_gemma":[0.0002592881,0.0001020084,0.00004608794,0.0001603442,0.0002025418,0.0001119693,0.0001430536,0.0001445617,0.00003249563],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00006014223,"about_ca_system_score_gemma":0.0000556383,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002533387,"about_ca_topic_score_gemma":0.0002224933,"domain_scores_codex":[0.9988306,0.0001009262,0.0003276349,0.0003439941,0.0001822997,0.0002145534],"domain_scores_gemma":[0.9990289,0.00004930712,0.00008435434,0.0004749698,0.0002507925,0.0001116556],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0001382832,0.0001055239,0.9013928,0.00005890794,0.0001680822,8.574618e-7,0.0000623466,0.000005165321,0.06843124,0.00003155401,0.00903769,0.02056759],"study_design_scores_gemma":[0.0005561266,0.0006763296,0.9577159,0.00004641913,0.000145072,0.00002124462,0.000009782255,0.01820222,0.006211034,0.0004793698,0.01586448,0.0000719715],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9893883,0.0007463658,0.006916359,0.001461847,0.000507752,0.0001891119,0.0002996131,0.00004466505,0.0004459698],"genre_scores_gemma":[0.9937357,0.002307006,0.001641654,0.0002389055,0.001423985,0.000009177127,0.000600977,0.00001325916,0.00002933452],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.06222021,"threshold_uncertainty_score":0.4159781,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04431531856062019,"score_gpt":0.3223620938385618,"score_spread":0.2780467752779416,"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."}}