{"id":"W4389028818","doi":"10.1101/2023.11.26.568742","title":"Learning sequence, structure, and function representations of proteins with language models","year":2023,"lang":"en","type":"preprint","venue":"bioRxiv (Cold Spring Harbor Laboratory)","topic":"Machine Learning in Bioinformatics","field":"Biochemistry, Genetics and Molecular Biology","cited_by":11,"is_retracted":false,"has_abstract":true,"ca_institutions":"Canadian Institute for Advanced Research","funders":"Flatiron Health; Division of Chemical, Bioengineering, Environmental, and Transport Systems; Samsung Advanced Institute of Technology; Samsung; National Institutes of Health; National Science Foundation","keywords":"Sequence (biology); Function (biology); Computer science; Computational biology; Natural language processing; Artificial intelligence; Biology; Evolutionary biology; Genetics","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":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0002438714,0.0002959518,0.000271957,0.0001506572,0.0001062913,0.00006991444,0.0002137453,0.0003534654,0.000006621863],"category_scores_gemma":[0.0001964467,0.000283021,0.00005384748,0.0002164077,0.0001336696,0.00001471668,0.0003860689,0.0005400131,0.000002743875],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002404565,"about_ca_system_score_gemma":0.0002333407,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00009945224,"about_ca_topic_score_gemma":0.000010655,"domain_scores_codex":[0.9985753,0.00009506394,0.0003332695,0.0005102749,0.0002431308,0.0002429195],"domain_scores_gemma":[0.9984958,0.00001741222,0.0004270601,0.0007016363,0.0002631966,0.0000948588],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.00006666028,0.0000200804,0.01361043,0.0005755351,0.0001977603,0.000006632713,0.00005409869,0.02579297,0.9592937,0.0003181031,0.00005848397,0.000005517661],"study_design_scores_gemma":[0.001269478,0.0008984758,0.08595781,0.00067234,0.0003284468,2.547516e-7,0.00009970536,0.0363646,0.8718711,0.00003054972,0.001105267,0.001401957],"study_design_candidate":"bench_or_experimental","study_design_consensus":"bench_or_experimental","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.983358,0.0002326072,0.01535762,0.00003700421,0.0001399361,0.0005486375,0.0001735006,0.0001290011,0.00002366435],"genre_scores_gemma":[0.98563,0.0001107166,0.01387629,0.00003134739,0.0001451714,0.00007195448,0.000009531235,0.00008685576,0.00003814423],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.08742262,"threshold_uncertainty_score":0.9999622,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01277411729885679,"score_gpt":0.2404532850849711,"score_spread":0.2276791677861143,"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."}}