{"id":"W4225839207","doi":"10.1093/bioinformatics/btac259","title":"PhyloPGM: boosting regulatory function prediction accuracy using evolutionary information","year":2022,"lang":"en","type":"article","venue":"Bioinformatics","topic":"Genomics and Chromatin Dynamics","field":"Biochemistry, Genetics and Molecular Biology","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University","funders":"","keywords":"Computational biology; Computer science; Boosting (machine learning); Context (archaeology); Probabilistic logic; Sequence (biology); Function (biology); Data mining; Extant taxon; Genomics; DNA binding site; Gene regulatory network; Machine learning; Biology; Genome; Gene; Artificial intelligence; Genetics; Gene expression; Evolutionary biology; Promoter","routes":{"ca_aff":true,"ca_fund":false,"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":[0.003599631,0.002706569,0.001815741,0.002409297,0.001218405,0.001133996,0.003097775,0.002043298,0.005320624],"category_scores_gemma":[0.009357446,0.0009290645,0.001555783,0.002083918,0.001072147,0.002141798,0.002805978,0.003112124,0.003135849],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0008853737,"about_ca_system_score_gemma":0.001765995,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004378402,"about_ca_topic_score_gemma":0.004294557,"domain_scores_codex":[0.9986922,0.0004195125,0.00004561179,0.0004620286,0.0002641466,0.0001165349],"domain_scores_gemma":[0.9977797,0.001289377,0.000166965,0.0003892849,0.0002535649,0.0001211077],"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":[0.002679226,0.0006142475,0.02597897,0.002004572,0.0009423129,0.0004906764,0.0003918106,0.5065345,0.03053376,0.007369217,0.09106396,0.3313968],"study_design_scores_gemma":[0.0001417153,0.0001324932,0.001569674,0.0000556026,0.00009348691,0.0001331908,0.00003696395,0.9774975,0.006889015,0.006922625,0.006490918,0.00003681257],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.2219834,0.006150721,0.5890898,0.002483799,0.0004307035,0.0003809934,0.01461369,0.1581765,0.006690375],"genre_scores_gemma":[0.4671389,0.001529711,0.48777,0.00121969,0.00025308,0.0006819516,0.031565,0.007044186,0.002797527],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.005320624,"threshold_uncertainty_score":0.01903689,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.008650630458668988,"score_gpt":0.2058492327359707,"score_spread":0.1971986022773017,"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."}}