{"id":"W4395082224","doi":"10.1002/prot.26694","title":"Exploiting protein language models for the precise classification of ion channels and ion transporters","year":2024,"lang":"en","type":"article","venue":"Proteins Structure Function and Bioinformatics","topic":"Machine Learning in Bioinformatics","field":"Biochemistry, Genetics and Molecular Biology","cited_by":8,"is_retracted":false,"has_abstract":true,"ca_institutions":"Concordia University","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Robustness (evolution); Computer science; Random forest; Generalization; Artificial intelligence; Machine learning; Biology; Mathematics","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00108768,0.001273707,0.0007690437,0.001150034,0.0004445992,0.001995237,0.001654985,0.00138994,0.001759671],"category_scores_gemma":[0.003771903,0.0003729426,0.001467946,0.0009160064,0.0005240863,0.003160332,0.001839449,0.002154898,0.00262188],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00100324,"about_ca_system_score_gemma":0.001587832,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003473319,"about_ca_topic_score_gemma":0.005385768,"domain_scores_codex":[0.9994294,0.0001194407,0.00004112593,0.0001988687,0.0001385571,0.00007258143],"domain_scores_gemma":[0.9991536,0.0003592381,0.0001089208,0.0001440599,0.0001723104,0.00006188899],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.000941758,0.0004994808,0.01445359,0.0007122939,0.0003260595,0.0008631119,0.0003876096,0.2926725,0.06624208,0.03686855,0.03951139,0.5465216],"study_design_scores_gemma":[0.00001800769,0.00005395703,0.0003736684,0.00002542467,0.00002058571,0.0001116953,0.00004584399,0.971262,0.00680175,0.01682741,0.00443687,0.00002285087],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.07624459,0.003875202,0.8931363,0.002217504,0.0004362999,0.0001757584,0.003937293,0.0136717,0.006305387],"genre_scores_gemma":[0.6018583,0.002036656,0.369316,0.002376953,0.0003004938,0.0004162909,0.0136926,0.0009820315,0.00902066],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.003473319,"threshold_uncertainty_score":0.007279038,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01184212033902125,"score_gpt":0.2373179589850951,"score_spread":0.2254758386460739,"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."}}