{"id":"W2012080711","doi":"10.1111/j.1755-0998.2009.02651.x","title":"Efficient algorithms for the discovery of DNA oligonucleotide barcodes from sequence databases","year":2009,"lang":"en","type":"article","venue":"Molecular Ecology Resources","topic":"Genomics and Phylogenetic Studies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":16,"is_retracted":false,"has_abstract":true,"ca_institutions":"Carleton University; Simon Fraser University; Agriculture and Agri-Food Canada","funders":"","keywords":"Biology; Oligonucleotide; Sequence (biology); Computational biology; Sequence database; DNA sequencing; DNA; Genetics; Database; Computer science; Gene","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.006416251,0.002820916,0.003222146,0.007926384,0.002141797,0.005913499,0.006465707,0.003699708,0.00451688],"category_scores_gemma":[0.02819175,0.002641585,0.002859131,0.009355324,0.001831791,0.00642489,0.004048154,0.004011798,0.005253912],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002203561,"about_ca_system_score_gemma":0.00358235,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002242367,"about_ca_topic_score_gemma":0.003904126,"domain_scores_codex":[0.9940044,0.001545603,0.0008127396,0.001245843,0.002077338,0.0003140292],"domain_scores_gemma":[0.9837172,0.01101543,0.001286236,0.001791983,0.001802475,0.0003866734],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0005661187,0.0004350431,0.003132055,0.001931142,0.0004377689,0.0004289733,0.0005378073,0.07263096,0.0168335,0.05753519,0.01903897,0.8264925],"study_design_scores_gemma":[0.0003070777,0.0001284069,0.0007775869,0.0002013389,0.0001274869,0.0008039919,0.0002500532,0.784433,0.01453351,0.1790444,0.01929009,0.0001030131],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.002234832,0.0003485693,0.9923124,0.0001448438,0.00003091096,0.0002346896,0.0003693028,0.003973661,0.0003508234],"genre_scores_gemma":[0.007436204,0.000252414,0.9898269,0.0000588572,0.0000223673,0.0003390682,0.001487777,0.0002185947,0.0003577173],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.007926384,"threshold_uncertainty_score":0.03393275,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01867341585266136,"score_gpt":0.2610099011311425,"score_spread":0.2423364852784811,"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."}}