{"id":"W2979310796","doi":"10.1111/2041-210x.13314","title":"MetaCurator: A hidden Markov model‐based toolkit for extracting and curating sequences from taxonomically‐informative genetic markers","year":2019,"lang":"en","type":"article","venue":"Methods in Ecology and Evolution","topic":"Environmental DNA in Biodiversity Studies","field":"Environmental Science","cited_by":54,"is_retracted":false,"has_abstract":true,"ca_institutions":"York University","funders":"Ohio State University","keywords":"Hidden Markov model; Pipeline (software); Sequence database; Biology; Sequence (biology); Computer science; Information retrieval; Artificial intelligence; Genetics; Programming language","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":[],"consensus_categories":[],"category_scores_codex":[0.0007010482,0.0001247509,0.0001971643,0.00003765231,0.0001865928,0.00001460033,0.00008637272,0.000117062,0.0001270581],"category_scores_gemma":[0.0001504415,0.000122259,0.00002919024,0.00005165425,0.000267046,0.0003352496,0.0001771383,0.0001212034,0.00001796159],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003041156,"about_ca_system_score_gemma":0.000005825547,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000151004,"about_ca_topic_score_gemma":0.0001206327,"domain_scores_codex":[0.9989872,0.0002104385,0.0002165732,0.0003030245,0.0000640911,0.0002186864],"domain_scores_gemma":[0.9989712,0.0007654083,0.0001311246,0.00008961332,0.000003066845,0.00003953623],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.00006517738,0.00001482533,0.9776576,0.00001146994,0.00001512156,4.446103e-7,0.0003258959,0.001686466,0.005002417,0.000009931083,0.00004737112,0.01516325],"study_design_scores_gemma":[0.0003856114,0.00007298688,0.7779077,0.00000679531,0.00002466633,0.000001228184,0.0008174328,0.2192645,0.0002735846,0.001074534,0.00005032391,0.0001206765],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"methods","genre_scores_codex":[0.9585583,0.0001218899,0.04004504,0.0001078914,0.00009758167,0.0004407884,0.00002383425,0.00001221301,0.0005925093],"genre_scores_gemma":[0.4836341,0.00002202405,0.5162007,0.0000767501,0.000004417275,0.00002986767,0.000003020361,0.000002608103,0.00002656518],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.4761556,"threshold_uncertainty_score":0.4985579,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02145089154093273,"score_gpt":0.2836678876402798,"score_spread":0.2622169960993471,"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."}}