{"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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.000200441,0.0001637125,0.0002777698,0.0000321538,0.00017774,0.00002317051,0.0002231174,0.00006072297,0.0006593618],"category_scores_gemma":[0.00003104521,0.0001664921,0.0001442376,0.00008769871,0.000372239,0.00008761276,0.0005543982,0.0000991861,0.00003367358],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002040144,"about_ca_system_score_gemma":0.00001527834,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001524962,"about_ca_topic_score_gemma":0.00003522096,"domain_scores_codex":[0.9987471,0.000113459,0.0002387512,0.0004053119,0.0002050533,0.0002903647],"domain_scores_gemma":[0.9994415,0.0001185206,0.0001455992,0.0002290775,0.000008200757,0.00005706836],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.000008201752,0.0002562979,0.8090233,0.00004038567,0.0001464137,0.00002300968,0.0002324924,0.003658772,0.1836397,0.0009415519,0.0008778077,0.001152126],"study_design_scores_gemma":[0.000594319,0.0001894947,0.9158234,0.00001351652,0.0001384058,0.000005848275,0.001689949,0.0002973156,0.05917572,0.0007902794,0.02101361,0.0002681661],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9885103,0.0005998919,0.005674246,0.00018458,0.0001357594,0.0003047064,0.00003491709,0.00001919974,0.004536384],"genre_scores_gemma":[0.9817462,0.00007597449,0.01786075,0.0001227038,0.00001212629,0.00006906242,0.00001575635,0.0000103493,0.00008710432],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.124464,"threshold_uncertainty_score":0.7219549,"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."}}