{"id":"W2168149629","doi":"10.1111/1755-0998.12240","title":"Rapid and accurate taxonomic classification of insect (class Insecta) cytochrome <i>c</i> oxidase subunit 1 (<scp>COI</scp>) <scp>DNA</scp> barcode sequences using a naïve Bayesian classifier","year":2014,"lang":"en","type":"article","venue":"Molecular Ecology Resources","topic":"Environmental DNA in Biodiversity Studies","field":"Environmental Science","cited_by":64,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of New Brunswick; University of Guelph; McMaster University","funders":"","keywords":"Biology; Taxonomic rank; Barcode; Cytochrome c oxidase subunit I; DNA barcoding; Supermatrix; Classifier (UML); Biological classification; Taxon; Artificial intelligence; Machine learning; Phylogenetics; Computational biology; Evolutionary biology; Computer science; Ecology; Genetics; Gene; Mathematics","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":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0005507344,0.0004580242,0.0005670925,0.0001928673,0.0005209975,0.00007022497,0.0006048246,0.0003577503,0.0001266472],"category_scores_gemma":[0.0006208181,0.0004668235,0.0001465347,0.0003559186,0.002159344,0.0002912132,0.0007856018,0.0003392146,0.0001535478],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003354645,"about_ca_system_score_gemma":0.00001577303,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002372302,"about_ca_topic_score_gemma":0.0002430712,"domain_scores_codex":[0.9968056,0.0006587626,0.0005185718,0.0009185107,0.0003994359,0.0006991233],"domain_scores_gemma":[0.9978137,0.0008669228,0.0005358013,0.0005277412,0.00001848229,0.0002373445],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"observational","study_design_scores_codex":[0.000007219126,0.0001019314,0.3427156,0.00003163493,0.00009670952,0.00002815624,0.0009769013,0.0004089527,0.6546878,0.00005163928,0.0007062831,0.0001872299],"study_design_scores_gemma":[0.00077574,0.0004525052,0.9154005,0.0000328865,0.0001407306,0.00004809095,0.002034376,0.004289746,0.06493425,0.0002661578,0.01143624,0.0001887706],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9930562,0.0002308928,0.0004534667,0.0001522864,0.00007810636,0.0005679922,0.00005294421,0.00006541366,0.005342649],"genre_scores_gemma":[0.9964757,0.0001231009,0.002503196,0.0005663879,0.00002884807,0.00005106386,0.00001975561,0.00003908045,0.0001928549],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.5897535,"threshold_uncertainty_score":0.9997783,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02302088286307289,"score_gpt":0.213235689719932,"score_spread":0.1902148068568591,"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."}}