{"id":"W2246311611","doi":"10.1371/journal.pone.0141603","title":"Computational Identification of MoRFs in Protein Sequences Using Hierarchical Application of Bayes Rule","year":2015,"lang":"en","type":"article","venue":"PLoS ONE","topic":"Machine Learning in Bioinformatics","field":"Biochemistry, Genetics and Molecular Biology","cited_by":50,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of British Columbia","funders":"Canadian Institutes of Health Research; Genome British Columbia; Michael Smith Health Research BC","keywords":"Bayes' theorem; Computer science; Identification (biology); Protein sequencing; Computational biology; Sequence (biology); Sequence alignment; Naive Bayes classifier; Artificial intelligence; Data mining; Bioinformatics; Peptide sequence; Biology; Bayesian probability; 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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.003950906,0.001039976,0.001828576,0.002826544,0.000834955,0.001091915,0.001681542,0.001281765,0.002582687],"category_scores_gemma":[0.009306688,0.0007812466,0.001709516,0.001185655,0.0006804466,0.001185221,0.0008686789,0.001624819,0.0009339934],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001081622,"about_ca_system_score_gemma":0.002594347,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01247592,"about_ca_topic_score_gemma":0.01317491,"domain_scores_codex":[0.9985941,0.0004686222,0.0001302705,0.0003170324,0.0003543346,0.0001355943],"domain_scores_gemma":[0.9904168,0.007822391,0.0004677147,0.0002970398,0.0007233549,0.0002727281],"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":[0.000562432,0.0002786743,0.02567402,0.0003474866,0.0004207003,0.0004253743,0.0001465852,0.8030598,0.002364276,0.003936813,0.003591308,0.1591924],"study_design_scores_gemma":[0.00001212996,0.00002178461,0.0003620884,0.00001135539,0.0000155289,0.00002142024,0.000006813857,0.9974226,0.0002360202,0.00178708,0.00009923321,0.000003939318],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.1980209,0.00163905,0.7898992,0.000632448,0.0001039161,0.0004138598,0.001215685,0.004916854,0.003158113],"genre_scores_gemma":[0.5270789,0.0004785257,0.4669659,0.0004118288,0.0001537102,0.0003674949,0.003147774,0.0002022408,0.001193638],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01247592,"threshold_uncertainty_score":0.02480656,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0328941551623359,"score_gpt":0.2779105196041278,"score_spread":0.2450163644417919,"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."}}