{"id":"W3216502886","doi":"10.3389/fgene.2021.768473","title":"Simple Matching Using QIIME 2 and RDP Reveals Misidentified Sequences and an Underrepresentation of Fungi in Reference Datasets","year":2021,"lang":"en","type":"article","venue":"Frontiers in Genetics","topic":"Mycorrhizal Fungi and Plant Interactions","field":"Agricultural and Biological Sciences","cited_by":8,"is_retracted":false,"has_abstract":true,"ca_institutions":"Western University","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Classifier (UML); Matching (statistics); Phylogenetic tree; Simple (philosophy); Computer science; Artificial intelligence; Pattern recognition (psychology); Biology; Mathematics; Genetics; Statistics","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":[0.00012059,0.00005989278,0.0001178959,0.00002149757,0.00004397929,0.00004353225,0.0000633561,0.00005285685,0.00007996566],"category_scores_gemma":[0.00003025427,0.00003383344,0.000008805933,0.0001554186,0.00001164174,0.0001188484,0.00004505129,0.00007168483,5.545711e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001020683,"about_ca_system_score_gemma":0.000008277694,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001629263,"about_ca_topic_score_gemma":0.002072031,"domain_scores_codex":[0.9993454,0.0001005005,0.0001822537,0.0001905671,0.00007678454,0.0001045429],"domain_scores_gemma":[0.9996676,0.0001720711,0.00005609586,0.00004606185,0.00001987107,0.00003834167],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.00001585963,0.00002658191,0.6222818,0.0000181898,0.000005450371,0.0000141353,0.000226767,0.0002109906,0.3589671,0.00007159743,0.0006301118,0.01753142],"study_design_scores_gemma":[0.0002606939,0.0001708179,0.8668665,0.0003183057,0.00005320896,0.000102483,0.01438745,0.02062304,0.02945443,0.06417533,0.003051923,0.0005357949],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9944288,0.004338785,0.000113791,0.00003298336,0.0001167371,0.00006773417,0.0008169452,0.000003579106,0.00008068432],"genre_scores_gemma":[0.9897559,0.003718096,0.005049246,0.0000321299,0.00002627798,0.000001194074,0.001397592,5.759037e-7,0.00001895465],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.3295127,"threshold_uncertainty_score":0.2462969,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05178709917892368,"score_gpt":0.3028447966177322,"score_spread":0.2510576974388086,"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."}}