{"id":"W3159353896","doi":"10.1111/1755-0998.13407","title":"Assessment of current taxonomic assignment strategies for metabarcoding eukaryotes","year":2021,"lang":"en","type":"article","venue":"Molecular Ecology Resources","topic":"Environmental DNA in Biodiversity Studies","field":"Environmental Science","cited_by":96,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Guelph; McGill University","funders":"Canada First Research Excellence Fund","keywords":"Barcode; Biology; Biodiversity; Taxonomic rank; Completeness (order theory); Set (abstract data type); Taxonomy (biology); Taxon; Selection (genetic algorithm); Computer science; Information retrieval; Data mining; Ecology; Machine learning; Mathematics","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.04775909,0.001424708,0.001004244,0.003686837,0.001599424,0.004521052,0.002991817,0.001730571,0.0007513264],"category_scores_gemma":[0.07708474,0.0008574663,0.001058322,0.002455283,0.001129034,0.004282305,0.002012866,0.001318573,0.001074653],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002119482,"about_ca_system_score_gemma":0.002223466,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004684656,"about_ca_topic_score_gemma":0.008509331,"domain_scores_codex":[0.9826787,0.007876883,0.00180391,0.002916225,0.004179083,0.0005451767],"domain_scores_gemma":[0.9636115,0.01619925,0.004655234,0.003572689,0.0109614,0.0009999672],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"observational","study_design_scores_codex":[0.002143237,0.0005195093,0.1304456,0.002941645,0.00117575,0.0001946169,0.003098919,0.04281317,0.1259308,0.006646979,0.004665107,0.6794246],"study_design_scores_gemma":[0.0002300719,0.002853613,0.1220078,0.0024093,0.001234796,0.001249041,0.003729403,0.5798962,0.2298808,0.01583181,0.04012962,0.0005474992],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.5824894,0.01739002,0.3833691,0.003221091,0.000340382,0.0007224231,0.001769371,0.00443521,0.006262972],"genre_scores_gemma":[0.4208887,0.002566481,0.571918,0.0004416078,0.00004505693,0.0003168901,0.002735895,0.0003865141,0.0007008844],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.04775909,"threshold_uncertainty_score":0.2525772,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02228986564391712,"score_gpt":0.2662165642978971,"score_spread":0.24392669865398,"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."}}