{"id":"W3207684739","doi":"10.1101/2020.09.14.297424","title":"leADS: improved metabolic pathway inference based on active dataset subsampling","year":2020,"lang":"en","type":"preprint","venue":"bioRxiv (Cold Spring Harbor Laboratory)","topic":"Machine Learning in Bioinformatics","field":"Biochemistry, Genetics and Molecular Biology","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"Genome British Columbia; University of British Columbia","funders":"Natural Sciences and Engineering Research Council of Canada; Genome British Columbia; Compute Canada; Genome Canada","keywords":"Inference; Computer science; Machine learning; Artificial intelligence; Class (philosophy); License; Feature (linguistics); MIT License; Set (abstract data type); Training set; Data mining; Programming language","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":[0.003321578,0.001231294,0.00165131,0.002763998,0.0007389935,0.002001459,0.00244527,0.001391183,0.004294916],"category_scores_gemma":[0.009497162,0.0005368243,0.00156478,0.001949685,0.0005471944,0.001953398,0.001756427,0.00163277,0.001938377],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009438192,"about_ca_system_score_gemma":0.002024376,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.008919178,"about_ca_topic_score_gemma":0.01799083,"domain_scores_codex":[0.9985673,0.0005322358,0.0000669605,0.0003945886,0.0003415615,0.0000973079],"domain_scores_gemma":[0.996241,0.001777804,0.0001851705,0.001102135,0.0005043277,0.0001895526],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00239983,0.001167557,0.02126097,0.0006504183,0.001113755,0.0004383337,0.0001862991,0.3355562,0.01418938,0.008664349,0.07253556,0.5418373],"study_design_scores_gemma":[0.00009614488,0.00005733993,0.0008858486,0.00001487145,0.00002959194,0.00003567624,0.00002694505,0.9868188,0.002815484,0.00666743,0.002536847,0.00001502502],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.1291157,0.001886596,0.7913235,0.001202555,0.0005798317,0.0004904724,0.01509434,0.0560814,0.004225718],"genre_scores_gemma":[0.4136183,0.0003497202,0.5345334,0.0007950859,0.0002952062,0.00039287,0.04457147,0.001413864,0.004030154],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.008919178,"threshold_uncertainty_score":0.01773453,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01607321600606083,"score_gpt":0.2517440644266227,"score_spread":0.2356708484205619,"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."}}