{"id":"W2338899252","doi":"10.1139/gen-2015-0184","title":"The effects of parameter choice on defining molecular operational taxonomic units and resulting ecological analyses of metabarcoding data","year":2016,"lang":"en","type":"article","venue":"Genome","topic":"Environmental DNA in Biodiversity Studies","field":"Environmental Science","cited_by":86,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University","funders":"","keywords":"Biology; Cluster analysis; Ecology; Taxonomic rank; Biodiversity; DNA barcoding; Simple (philosophy); Evolutionary biology; Computer science; Taxon; Machine learning","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":true,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.06771549,0.002072502,0.001544215,0.001644774,0.001393314,0.003413128,0.001522377,0.001964764,0.0006420987],"category_scores_gemma":[0.1962093,0.000847043,0.001699026,0.002402732,0.002162417,0.003212748,0.002094164,0.003519013,0.0004055071],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001339566,"about_ca_system_score_gemma":0.001040719,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002486656,"about_ca_topic_score_gemma":0.003302716,"domain_scores_codex":[0.9196321,0.06221653,0.006376848,0.00645622,0.004306875,0.001011388],"domain_scores_gemma":[0.7092076,0.2640466,0.006621598,0.01459663,0.004689869,0.0008378471],"domain_codex":null,"domain_gemma":"methods","domain_candidate":"methods","domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.008139186,0.001531978,0.3637055,0.003407374,0.006188115,0.001123972,0.003800138,0.2092614,0.1863019,0.008174315,0.006498905,0.2018673],"study_design_scores_gemma":[0.0005755278,0.003676205,0.3375645,0.001295493,0.002686513,0.002311273,0.002599646,0.3383562,0.2437804,0.0440387,0.02183108,0.001284505],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8187231,0.004143616,0.1669869,0.002697983,0.0004835317,0.0005504281,0.001946118,0.00144164,0.003026795],"genre_scores_gemma":[0.8940494,0.0007578011,0.1002803,0.001155303,0.00007294011,0.0007233765,0.001737138,0.00094442,0.0002793007],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.9322845,"threshold_uncertainty_score":0.358118,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06556737303967468,"score_gpt":0.271968952736585,"score_spread":0.2064015796969103,"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."}}