{"id":"W4297991272","doi":"10.1371/journal.pone.0274260","title":"MetaWorks: A flexible, scalable bioinformatic pipeline for high-throughput multi-marker biodiversity assessments","year":2022,"lang":"en","type":"article","venue":"PLoS ONE","topic":"Environmental DNA in Biodiversity Studies","field":"Environmental Science","cited_by":46,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Guelph","funders":"Government of Canada; Ontario Genomics; Genome Canada","keywords":"Computer science; Scalability; Pipeline (software); Workflow; Computational biology; Biology; Classifier (UML); Data science; Database; Artificial intelligence","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":["insufficient_payload"],"category_scores_codex":[0.0004090104,0.0002401972,0.000338274,0.00006558078,0.001243749,0.0000364102,0.0005295035,0.00005296856,0.006288675],"category_scores_gemma":[0.00004347535,0.0002564298,0.0001052768,0.0003355406,0.0002652582,0.0004725841,0.002165785,0.0002136921,0.001635746],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007737711,"about_ca_system_score_gemma":0.000005101507,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0003616354,"about_ca_topic_score_gemma":0.00001667736,"domain_scores_codex":[0.9978429,0.00007319244,0.0002840844,0.0004560538,0.000859507,0.000484279],"domain_scores_gemma":[0.9992279,0.00009514377,0.0001715689,0.0003778778,0.000009525262,0.00011803],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0002406224,0.00702186,0.9088168,0.0001521222,0.0007684148,0.00001087331,0.0007047133,0.001071367,0.007436701,0.00001873141,0.07204267,0.001715179],"study_design_scores_gemma":[0.01344452,0.001954354,0.8534849,0.0001158253,0.002815112,0.000008805705,0.00636869,0.03820119,0.04429991,0.00056968,0.03570648,0.003030473],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9901775,0.0001253067,0.003221575,0.000867938,0.0001982522,0.001791419,0.001218028,0.0002148781,0.002185166],"genre_scores_gemma":[0.5787749,0.0002191271,0.401147,0.002803129,0.00004302644,0.0004442767,0.0005092036,0.00003503799,0.01602431],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.4114026,"threshold_uncertainty_score":0.9999888,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05554969787539162,"score_gpt":0.2388530373279584,"score_spread":0.1833033394525668,"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."}}