{"id":"W4396655022","doi":"10.1098/rstb.2023.0124","title":"PROTAX-GPU: a scalable probabilistic taxonomic classification system for DNA barcodes","year":2024,"lang":"en","type":"article","venue":"Philosophical Transactions of the Royal Society B Biological Sciences","topic":"Species Distribution and Climate Change","field":"Environmental Science","cited_by":6,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Guelph; Vector Institute; University of Toronto","funders":"H2020 European Research Council; Canada Research Chairs; Canada First Research Excellence Fund; Natural Sciences and Engineering Research Council of Canada; European Commission; Government of Canada; Canadian Institute for Advanced Research","keywords":"Computer science; Scalability; Probabilistic logic; Data mining; Identification (biology); Python (programming language); Taxonomic rank; Data science; Machine learning; Information retrieval; Database; Artificial intelligence; Taxon; Ecology; 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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0007695158,0.001507187,0.001044684,0.001315789,0.0008840426,0.001870033,0.004609944,0.0009915291,0.009786353],"category_scores_gemma":[0.003768961,0.000860373,0.0009970564,0.001963349,0.0006634788,0.002096134,0.002559523,0.0019106,0.005998502],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001260581,"about_ca_system_score_gemma":0.002070316,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.009340686,"about_ca_topic_score_gemma":0.009192525,"domain_scores_codex":[0.9992507,0.00008342842,0.00006468835,0.00020116,0.0003222386,0.00007780219],"domain_scores_gemma":[0.99921,0.000170701,0.00007909679,0.0001459222,0.0003017815,0.00009241356],"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":[0.001509892,0.0002332479,0.006173686,0.001113838,0.0003669946,0.0005263184,0.0007419912,0.04695445,0.04029458,0.02413016,0.3084602,0.5694946],"study_design_scores_gemma":[0.0002746471,0.0001532678,0.002576925,0.0001092297,0.00007221521,0.000324034,0.0001081898,0.8101285,0.02813239,0.02469447,0.1332698,0.0001564422],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.01892552,0.001285685,0.7410858,0.0005540318,0.00052519,0.0003255876,0.006518784,0.2223117,0.008467795],"genre_scores_gemma":[0.1208982,0.0008076652,0.8378487,0.0007506104,0.0001158205,0.0009544183,0.0206631,0.01046258,0.007498843],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.009786353,"threshold_uncertainty_score":0.03273857,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.09641066618165976,"score_gpt":0.2779848889428748,"score_spread":0.181574222761215,"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."}}