{"id":"W4319265453","doi":"10.1101/2023.02.03.527046","title":"Data-driven identification of major axes of functional variation in bacteria","year":2023,"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":"Université de Montréal; Espace pour la vie; Université du Québec à Montréal","funders":"Natural Sciences and Engineering Research Council of Canada; Canada Research Chairs","keywords":"Biology; Metagenomics; Evolutionary biology; Phylogenetic tree; Taxonomic rank; Trait; Phylogenetics; Tree of life (biology); Bacterial genome size; Ecology; Ecosystem; Genome; Taxon; Genetics; Gene","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":[0.00178237,0.0003920947,0.000506521,0.001602395,0.0003823242,0.001352848,0.0003288547,0.0003181739,0.0004770133],"category_scores_gemma":[0.006244016,0.0002771956,0.0008924471,0.001325934,0.0005663049,0.0005247288,0.0007892355,0.0006800328,0.000171373],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005608968,"about_ca_system_score_gemma":0.0005791815,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001092023,"about_ca_topic_score_gemma":0.001704061,"domain_scores_codex":[0.9989084,0.0003838354,0.00007195798,0.0003430918,0.0002165628,0.00007611986],"domain_scores_gemma":[0.9963239,0.002091691,0.0006277302,0.0003817977,0.0004210077,0.000153789],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0007895751,0.0002617895,0.6339694,0.000647684,0.0009921782,0.0003108038,0.0003265643,0.0411593,0.2578542,0.005150161,0.001054233,0.05748414],"study_design_scores_gemma":[0.00005631467,0.0003016799,0.5861466,0.00006405604,0.0002474348,0.0003359487,0.0003688273,0.3554684,0.03908361,0.01502567,0.002812112,0.0000893309],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9434855,0.0004026944,0.05018831,0.0002237904,0.00001739038,0.00005252158,0.004606013,0.0003137498,0.0007098756],"genre_scores_gemma":[0.9691598,0.00008111854,0.02620984,0.00005732779,0.000009966352,0.00005894755,0.004283226,0.00004749261,0.00009227708],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.00178237,"threshold_uncertainty_score":0.009426177,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02825120715009071,"score_gpt":0.2370100415238109,"score_spread":0.2087588343737202,"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."}}