{"id":"W4383092580","doi":"10.3389/fevo.2023.1190549","title":"Taxonomic distribution of opsin families inferred from UniProt Reference Proteomes and a suite of opsin-specific hidden Markov models","year":2023,"lang":"en","type":"article","venue":"Frontiers in Ecology and Evolution","topic":"Photoreceptor and optogenetics research","field":"Neuroscience","cited_by":5,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Victoria","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Opsin; Biology; UniProt; Evolutionary biology; Proteome; Hidden Markov model; Rhodopsin; Computational biology; Artificial intelligence; Genetics; Computer science; Gene; Retinal","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":[0.0002755821,0.0001068258,0.0002545844,0.0002456896,0.00008554976,0.00001174689,0.0001493977,0.0001659153,0.00001047989],"category_scores_gemma":[0.0001663382,0.0001088629,0.0000213054,0.0003547937,0.0005378276,0.0001617653,0.00014354,0.0001587184,0.000002969227],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001126071,"about_ca_system_score_gemma":0.00007092088,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000143103,"about_ca_topic_score_gemma":0.00005646203,"domain_scores_codex":[0.9987946,0.0002042233,0.0002745366,0.000334557,0.0001298683,0.0002622053],"domain_scores_gemma":[0.9995193,0.000128777,0.0001015058,0.0001557688,0.0000438094,0.00005080198],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0006988868,0.0001180813,0.7527499,0.00008786877,0.00002132723,0.00000520388,0.0006113565,0.0003766976,0.2318318,0.00200697,0.002619094,0.008872874],"study_design_scores_gemma":[0.0007269635,0.0001637432,0.9219248,0.00002786083,0.000006691304,0.000001419156,0.0003990315,0.02367189,0.03979458,0.0129164,0.0002424963,0.0001241106],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9964482,0.0003331135,0.001872703,0.00004333775,0.0002550597,0.0005270738,0.0003068175,0.00001950203,0.0001941881],"genre_scores_gemma":[0.9961987,0.002801074,0.0007657566,0.000004663583,0.00001341401,0.00008250352,0.00006807297,0.000006851175,0.00005901637],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.1920372,"threshold_uncertainty_score":0.4439299,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0416253657697335,"score_gpt":0.2614366393900854,"score_spread":0.2198112736203519,"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."}}