{"id":"W4388622124","doi":"10.1093/nar/gkad1050","title":"OpenProt 2.0 builds a path to the functional characterization of alternative proteins","year":2023,"lang":"en","type":"article","venue":"Nucleic Acids Research","topic":"Advanced Proteomics Techniques and Applications","field":"Chemistry","cited_by":58,"is_retracted":false,"has_abstract":true,"ca_institutions":"Centre Hospitalier Universitaire de Sherbrooke; Université de Sherbrooke","funders":"Fonds de Recherche du Québec - Santé; Alliance de recherche numérique du Canada; Canada Research Chairs","keywords":"Pipeline (software); Proteome; Biology; Computational biology; Computer science; Visualization; Resource (disambiguation); Open reading frame; Proteogenomics; Identification (biology); Interface (matter); Set (abstract data type); Genome; Artificial intelligence; Bioinformatics; Genomics; Gene; Peptide sequence; Genetics; Programming language","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0006050365,0.00008712998,0.0001040868,0.0001125417,0.000283721,0.00003984753,0.0005002947,0.00007151257,0.0005865627],"category_scores_gemma":[0.0001517738,0.00006661243,0.00004203208,0.0008906144,0.0001081574,0.00008406449,0.0003898763,0.0003618184,0.0002887844],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00007789029,"about_ca_system_score_gemma":0.00006951314,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00005186893,"about_ca_topic_score_gemma":0.000002987995,"domain_scores_codex":[0.998625,0.00003991877,0.0001982417,0.0002796051,0.0005618634,0.0002953962],"domain_scores_gemma":[0.9989746,0.00007330874,0.00005922938,0.000487856,0.0003312996,0.0000737294],"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":[0.00005187961,0.00004581129,0.0003899737,0.00004047266,0.00001153209,0.000001349161,0.0002725465,0.00002734176,0.9821979,0.007476285,0.001154565,0.008330382],"study_design_scores_gemma":[0.0002229115,0.00008622857,0.004692818,0.00008618137,0.000003001955,0.00000268001,0.0002115564,0.002619394,0.8917007,0.006660719,0.09357598,0.0001378603],"study_design_candidate":"bench_or_experimental","study_design_consensus":"bench_or_experimental","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9611761,0.000003972989,0.02729558,0.004524098,0.00002165992,0.001192154,0.000176392,0.0001584089,0.00545166],"genre_scores_gemma":[0.987411,0.00005513387,0.003619046,0.00006291408,0.0002479533,0.00283629,0.0001397655,0.00003678261,0.005591176],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.09242141,"threshold_uncertainty_score":0.642245,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06630731027716714,"score_gpt":0.36771572816234,"score_spread":0.3014084178851729,"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."}}