{"id":"W3138735990","doi":"10.1111/1755-0998.13384","title":"Debar: A sequence‐by‐sequence denoiser for COI‐5P DNA barcode data","year":2021,"lang":"en","type":"article","venue":"Molecular Ecology Resources","topic":"Environmental DNA in Biodiversity Studies","field":"Environmental Science","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Guelph","funders":"Ontario Ministry of Economic Development, Job Creation and Trade","keywords":"Barcode; Indel; Biology; DNA barcoding; In silico; Computational biology; Hidden Markov model; Sequence (biology); DNA sequencing; Word error rate; Genetics; Evolutionary biology; Artificial intelligence; Computer science; DNA; Gene; Single-nucleotide polymorphism","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","insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.0002102039,0.000244205,0.0002748467,0.0000245563,0.0004332777,0.00004289289,0.001066497,0.0001620151,0.001467425],"category_scores_gemma":[0.0002115375,0.000257477,0.00008356666,0.0001690615,0.0009066535,0.0001857566,0.00224163,0.0001519873,0.0005978083],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002381733,"about_ca_system_score_gemma":0.00001077631,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00008550518,"about_ca_topic_score_gemma":0.0002914523,"domain_scores_codex":[0.9977859,0.000148508,0.0002296875,0.000970913,0.00030038,0.0005645723],"domain_scores_gemma":[0.9987124,0.0001449529,0.00009800106,0.0008928687,0.00001017954,0.0001415624],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0000296283,0.0001574755,0.2692165,0.00001947084,0.0001450304,0.0003776361,0.0003689933,0.0002518323,0.6890103,0.00002616141,0.03920988,0.001187183],"study_design_scores_gemma":[0.001901982,0.0004474835,0.1893253,0.00002933364,0.0003943025,0.0002016793,0.001397478,0.001205643,0.2343198,0.001259736,0.5681518,0.001365459],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9938128,0.0005928922,0.0005850532,0.001944124,0.000103663,0.0003738273,0.0006819513,0.00005783913,0.001847869],"genre_scores_gemma":[0.9716472,0.0002136339,0.02002933,0.004935744,0.00003215917,0.00008639324,0.0004781242,0.00003962062,0.002537787],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.5289419,"threshold_uncertainty_score":0.9999877,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03547885858086927,"score_gpt":0.2606308441238946,"score_spread":0.2251519855430254,"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."}}