{"id":"W4414169396","doi":"10.1111/1755-0998.70043","title":"<scp>BOLDistilled</scp> : Automated Construction of Comprehensive but Compact <scp>DNA</scp> Barcode Reference Libraries","year":2025,"lang":"en","type":"article","venue":"Molecular Ecology Resources","topic":"Environmental DNA in Biodiversity Studies","field":"Environmental Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Guelph","funders":"Canada Foundation for Innovation; Government of Canada; Ontario Genomics; Genome Canada","keywords":"Barcode; DNA barcoding; Set (abstract data type); DNA sequencing; Construct (python library); Identification (biology); Mitochondrial DNA; Sequence (biology)","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":["metaepi_narrow","sts"],"consensus_categories":[],"category_scores_codex":[0.0001002669,0.0003577264,0.0005524149,0.0001534331,0.0004389994,0.0000453386,0.0006079281,0.0002795942,0.0001579961],"category_scores_gemma":[0.0002565103,0.0003657657,0.0001330177,0.0004072095,0.002894293,0.0001670439,0.0009629692,0.0002782563,0.0003101581],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000189208,"about_ca_system_score_gemma":0.00001470014,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002092186,"about_ca_topic_score_gemma":0.00004132885,"domain_scores_codex":[0.9977559,0.0002924463,0.0004218713,0.0006418257,0.0003433082,0.0005446369],"domain_scores_gemma":[0.9984561,0.0006678893,0.0002905413,0.0004400195,0.00002175344,0.0001237509],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.00001334892,0.0001988775,0.8574699,0.00006627183,0.000319735,0.00007892168,0.001245532,0.002399982,0.1117924,0.0005190073,0.02570367,0.000192375],"study_design_scores_gemma":[0.0006141011,0.0001762021,0.8827226,0.00003003016,0.00009896758,0.0000180988,0.00413041,0.0004499383,0.06313267,0.0005002542,0.04804694,0.00007985165],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9699191,0.0004823873,0.00007566003,0.0001623149,0.0001755086,0.0003743533,0.0001042724,0.0002693419,0.02843701],"genre_scores_gemma":[0.9935378,0.0001037583,0.003957427,0.000586833,0.000009243585,0.00001580155,0.00004785092,0.00001870533,0.001722599],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.04865973,"threshold_uncertainty_score":0.9998794,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.009412319452694063,"score_gpt":0.2132295893258004,"score_spread":0.2038172698731064,"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."}}