{"id":"W4205231340","doi":"10.1101/2022.01.12.475159","title":"MT-MAG: Accurate and interpretable machine learning for complete or partial taxonomic assignments of metagenome-assembled genomes","year":2022,"lang":"en","type":"preprint","venue":"bioRxiv (Cold Spring Harbor Laboratory)","topic":"Genomics and Phylogenetic Studies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"Vector Institute; Government of Canada; Canadian Institute for Advanced Research","keywords":"Metagenomics; Weighting; Genome; Taxonomic rank; Artificial intelligence; Biological classification; Pattern recognition (psychology); Feature (linguistics); Computer science; Biology; Software; Machine learning; Computational biology; Genetics; Evolutionary biology; Gene; Physics; Ecology","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":[0.002704684,0.002059836,0.001042645,0.002656502,0.0006337047,0.001987734,0.002486195,0.001511364,0.004616228],"category_scores_gemma":[0.0113738,0.000660716,0.00194855,0.001410228,0.0005770752,0.002481503,0.002144151,0.001936537,0.002873503],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007567075,"about_ca_system_score_gemma":0.0009066344,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001355833,"about_ca_topic_score_gemma":0.001820311,"domain_scores_codex":[0.9985245,0.0003449975,0.000121055,0.0004804177,0.0004362481,0.00009279936],"domain_scores_gemma":[0.9966722,0.00168652,0.0003677049,0.000611794,0.0004953598,0.0001664959],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.002213638,0.0003867034,0.02329689,0.001477334,0.0009037412,0.0004866986,0.0006364714,0.1444469,0.07434706,0.01480653,0.08777426,0.6492237],"study_design_scores_gemma":[0.00006232259,0.00008813066,0.001919021,0.00005293094,0.00003612339,0.00008730683,0.00005255401,0.9470004,0.02591378,0.01321428,0.0115196,0.00005364914],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.04164992,0.0005960611,0.7582513,0.0003929342,0.0002474782,0.0001154058,0.007599807,0.1892848,0.001862291],"genre_scores_gemma":[0.2277988,0.0002121296,0.7512887,0.0003891452,0.0001054441,0.0003546429,0.01355806,0.004782333,0.001510805],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.004616228,"threshold_uncertainty_score":0.01544285,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02778517443065139,"score_gpt":0.2436182446249829,"score_spread":0.2158330701943315,"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."}}