{"id":"W2952043149","doi":"10.1371/journal.pone.0200177","title":"Over 2.5 million COI sequences in GenBank and growing","year":2018,"lang":"en","type":"article","venue":"PLoS ONE","topic":"Environmental DNA in Biodiversity Studies","field":"Environmental Science","cited_by":176,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Guelph; Natural Resources Canada","funders":"","keywords":"GenBank; Barcode; Usability; Annotation; Popularity; Biology; Database; Geography; Computer science; Evolutionary biology; Information retrieval; Bioinformatics; Genetics; Gene","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":["insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.00008524714,0.00007583473,0.00009294771,0.00002121835,0.0001180382,0.0000104272,0.00009176787,0.00003351172,0.0009790905],"category_scores_gemma":[0.00001358631,0.00007410433,0.00001043828,0.0000903999,0.0004974105,0.0002429657,0.0002849515,0.00005474465,0.0006242409],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001145924,"about_ca_system_score_gemma":4.818463e-7,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0003721168,"about_ca_topic_score_gemma":0.0001409998,"domain_scores_codex":[0.9992917,0.00001903525,0.00007751941,0.0002183574,0.0002320532,0.0001613114],"domain_scores_gemma":[0.9998246,0.00002043064,0.00002289915,0.00009253038,0.000001166472,0.00003841244],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.000004960064,0.0001136034,0.8839281,0.000005216235,0.00001205063,0.000004268691,0.0004718678,0.000001741817,0.1150625,0.000003756553,0.0002488895,0.0001429938],"study_design_scores_gemma":[0.0001822227,0.00008898928,0.9640775,0.00003191223,0.00001685002,6.326774e-7,0.0001692554,0.0001218987,0.0347853,0.0001300361,0.000277811,0.0001175896],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9941586,0.00007540522,0.000001005791,0.0002708794,0.0000228448,0.0001024164,0.000004662842,0.00001568126,0.005348519],"genre_scores_gemma":[0.9958895,0.0002609425,0.003092869,0.0003496434,0.00002647996,0.000003901991,0.000001302416,0.00000404805,0.0003712401],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.08027725,"threshold_uncertainty_score":0.9999341,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03230627986650756,"score_gpt":0.2040242507341725,"score_spread":0.1717179708676649,"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."}}