{"id":"W4413491758","doi":"10.1093/bioinformatics/btaf458","title":"An interpretable alphabet for local protein structure search based on amino acid neighborhoods","year":2025,"lang":"en","type":"article","venue":"Bioinformatics","topic":"Protein Structure and Dynamics","field":"Biochemistry, Genetics and Molecular Biology","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Canadian Aeronautics and Space Institute; Wellcome Trust","keywords":"Alphabet; Computer science; Amino acid; Computational biology; Artificial intelligence; Biology; Genetics","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001168843,0.0005008598,0.0008306899,0.00135739,0.0006452039,0.001424756,0.001820414,0.0008745465,0.00610383],"category_scores_gemma":[0.007208938,0.0003510574,0.0005488707,0.001421253,0.0007676438,0.001625339,0.001578466,0.000963629,0.001995666],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001051111,"about_ca_system_score_gemma":0.001363478,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001275872,"about_ca_topic_score_gemma":0.00236506,"domain_scores_codex":[0.9989573,0.0002756902,0.0001506972,0.0002085905,0.0003322225,0.00007553147],"domain_scores_gemma":[0.9973532,0.001236998,0.0002041247,0.0005758553,0.000456394,0.0001734389],"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":[0.002198726,0.0005715973,0.01461679,0.00137077,0.00008395492,0.000756259,0.0007420172,0.1914025,0.06858584,0.1351242,0.04517606,0.5393714],"study_design_scores_gemma":[0.0002111986,0.0003020554,0.00147038,0.0001648798,0.00004193748,0.0004564506,0.0001721816,0.8173048,0.02693469,0.1155423,0.03732896,0.00007015635],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.07335699,0.0008363401,0.8879939,0.0005037856,0.000231965,0.0002405035,0.003865479,0.01949167,0.0134793],"genre_scores_gemma":[0.3681324,0.0003349505,0.6217182,0.0003334897,0.00005504021,0.0004735891,0.004719842,0.001601831,0.002630579],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.00610383,"threshold_uncertainty_score":0.02041942,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.004914227286936982,"score_gpt":0.2577339499135217,"score_spread":0.2528197226265847,"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."}}