{"id":"W4414398266","doi":"10.33137/codex.v1i1.45680","title":"Predicting Protein Functions: A Deep Learning Approach to Unraveling Biological Complexity","year":2025,"lang":"en","type":"article","venue":"Journal of Computing Data and Exploration","topic":"Computational Drug Discovery Methods","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"","keywords":"Deep learning; Protein sequencing; Field (mathematics); Sequence (biology); Feature (linguistics); Annotation; Biological data; Genomics","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":[0.0006013603,0.0006530449,0.0005108349,0.0009115434,0.0002990194,0.0008610151,0.001005034,0.0007807174,0.001136162],"category_scores_gemma":[0.001428446,0.0003262926,0.0007302974,0.0006485432,0.0006916324,0.00155737,0.0009791837,0.001713649,0.0003075134],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009882395,"about_ca_system_score_gemma":0.0009195704,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003533379,"about_ca_topic_score_gemma":0.00530266,"domain_scores_codex":[0.9998267,0.00004205897,0.000009633774,0.00004961088,0.00004854069,0.0000234575],"domain_scores_gemma":[0.999589,0.0002077758,0.00004445295,0.00005683082,0.00006815603,0.00003376334],"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":[0.0001529319,0.0002950985,0.005240217,0.0002006978,0.0001524496,0.0001524569,0.0001797612,0.5233296,0.02220479,0.0518042,0.005503135,0.3907846],"study_design_scores_gemma":[0.000002943876,0.00001609285,0.0001787118,0.000008402805,0.00000708448,0.00001299094,0.00000923838,0.9779163,0.001297039,0.01984526,0.0007017249,0.000004176982],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.0408644,0.001058609,0.9536008,0.001367055,0.00006315659,0.00003673248,0.0003650078,0.0007411158,0.001902986],"genre_scores_gemma":[0.6022662,0.002227944,0.3887365,0.0005654988,0.0001071439,0.000113654,0.001206153,0.0001264225,0.004650589],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.003533379,"threshold_uncertainty_score":0.00717026,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1447900144032077,"score_gpt":0.3506819999506972,"score_spread":0.2058919855474894,"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."}}