{"id":"W4411705537","doi":"10.1002/edn3.70125","title":"Automating the Curation of <scp>DNA</scp> Barcode Databases for Vascular Plants","year":2025,"lang":"en","type":"article","venue":"Environmental DNA","topic":"Environmental DNA in Biodiversity Studies","field":"Environmental Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Guelph","funders":"Gordon and Betty Moore Foundation","keywords":"Barcode; Database; Computer science; Data curation; World Wide Web; Operating system","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":[],"consensus_categories":[],"category_scores_codex":[0.0002546756,0.0002447398,0.0002272158,0.0000388847,0.000537894,0.00001875086,0.0004449316,0.00006035234,0.0003104239],"category_scores_gemma":[0.00008519855,0.0002023687,0.0001511139,0.0001010223,0.000673252,0.0003689321,0.0007915689,0.0001127586,0.0002849456],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002812472,"about_ca_system_score_gemma":0.000002863858,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00006774069,"about_ca_topic_score_gemma":0.00001731558,"domain_scores_codex":[0.998352,0.00007865446,0.0003307756,0.0004509433,0.0004344942,0.0003531515],"domain_scores_gemma":[0.9989141,0.0003904353,0.0001670201,0.0004719113,9.114367e-7,0.00005560484],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.00001375263,0.0003947522,0.6515006,0.00005780076,0.0001920583,0.000003059101,0.001149837,0.000884844,0.3206717,0.0001882398,0.01829432,0.006648971],"study_design_scores_gemma":[0.0006192062,0.00006696009,0.859938,0.00003855362,0.00009273132,0.000001917491,0.002044165,0.0006492975,0.103998,0.0001713643,0.03226548,0.000114341],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9937034,0.000304152,0.001490722,0.0001204014,0.0001695991,0.0007847283,0.0004441936,0.00004268421,0.00294009],"genre_scores_gemma":[0.9928686,0.0003023501,0.004951968,0.0003551381,0.00002779589,0.0000822081,0.0001877911,0.00001792629,0.001206225],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.2166737,"threshold_uncertainty_score":0.8252358,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01395009620961197,"score_gpt":0.2246438049989975,"score_spread":0.2106937087893856,"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."}}