{"id":"W4404018000","doi":"10.1101/2024.10.30.621013","title":"MethylGPT: a foundation model for the DNA methylome","year":2024,"lang":"en","type":"preprint","venue":"bioRxiv (Cold Spring Harbor Laboratory)","topic":"Epigenetics and DNA Methylation","field":"Biochemistry, Genetics and Molecular Biology","cited_by":25,"is_retracted":false,"has_abstract":true,"ca_institutions":"Vector Institute; University of Toronto; University Health Network","funders":"Medical Research Council; Wellcome Trust; Royal College of Physicians of Edinburgh; Scottish Funding Council; Scottish Government; Chief Scientist Office, Scottish Government Health and Social Care Directorate; National Alliance for Research on Schizophrenia and Depression","keywords":"Foundation (evidence); DNA methylation; Computer science; Computational biology; Biology; Genetics; Political science; Gene; Law; Gene expression","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":[0.0007300684,0.0006984576,0.0005188222,0.0005310694,0.000301679,0.0005461484,0.00132012,0.0008820482,0.002177504],"category_scores_gemma":[0.002706931,0.0004726386,0.001259181,0.000430786,0.0004652377,0.0006965236,0.0009477374,0.001389656,0.000924312],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007500216,"about_ca_system_score_gemma":0.001487579,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01038408,"about_ca_topic_score_gemma":0.01488315,"domain_scores_codex":[0.9998362,0.00004356079,0.000007305157,0.0000618241,0.00002827675,0.00002290624],"domain_scores_gemma":[0.999613,0.0002065411,0.00003262023,0.00005288539,0.00006838965,0.00002655944],"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":[0.0003698642,0.00007602778,0.01017748,0.0001355235,0.0002422424,0.0002415441,0.00009730846,0.8586804,0.009188456,0.009341154,0.00886712,0.1025828],"study_design_scores_gemma":[0.00001168301,0.00002258994,0.0003271851,0.000005344138,0.00001358617,0.00003396243,0.000003739687,0.9947619,0.0009083965,0.00317442,0.0007316616,0.000005600288],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.08379457,0.0004533699,0.9037589,0.0005795748,0.00009999863,0.0001039988,0.003381696,0.005510263,0.002317561],"genre_scores_gemma":[0.7727935,0.0005453303,0.2086852,0.0006411318,0.00008619768,0.0005278496,0.009472043,0.0006058455,0.006642812],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01038408,"threshold_uncertainty_score":0.02064729,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02216721304951737,"score_gpt":0.2636990197036043,"score_spread":0.2415318066540869,"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."}}