{"id":"W4322493887","doi":"10.1128/msystems.01284-22","title":"A Schema for Digitized Surface Swab Site Metadata in Open-Source DNA Sequence Databases","year":2023,"lang":"en","type":"article","venue":"mSystems","topic":"Genomics and Phylogenetic Studies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":16,"is_retracted":false,"has_abstract":true,"ca_institutions":"Simon Fraser University","funders":"National Institute of Food and Agriculture; U.S. Food and Drug Administration; U.S. Department of Agriculture","keywords":"Metadata; Computer science; Information retrieval; Metadata repository; World Wide Web; Database","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0004694283,0.0001603572,0.000267171,0.00003400664,0.00008764041,0.0001047859,0.0004084667,0.00005795456,0.000002324888],"category_scores_gemma":[0.000142802,0.0001506612,0.00006107147,0.0001825285,0.00004105036,0.000003907143,0.0006346346,0.00003784167,0.00002299097],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001327165,"about_ca_system_score_gemma":0.00006167468,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0003788064,"about_ca_topic_score_gemma":0.0004868735,"domain_scores_codex":[0.9988077,0.00005134854,0.0002624313,0.0004739465,0.00008902441,0.000315492],"domain_scores_gemma":[0.9992121,0.00005722256,0.00008073724,0.0005430339,0.00005011571,0.00005672433],"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":[0.00006557059,0.00002418089,0.007400978,0.00008175957,0.00007168,0.000004496022,0.0000728739,0.0004278687,0.9853601,0.0001865717,0.006099659,0.0002043022],"study_design_scores_gemma":[0.001771641,0.0001421265,0.001886367,0.00007787246,0.00002688999,0.00001340317,0.0004382269,0.001057791,0.1856147,0.00005620412,0.808488,0.0004268301],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9957262,0.001194122,0.0007037705,0.0001382783,0.0001875674,0.0007169711,0.0008820184,0.00001451876,0.00043661],"genre_scores_gemma":[0.9924403,0.0002332494,0.0007890364,0.00009935338,0.0001232067,0.0001372659,0.000849493,0.00003288633,0.005295209],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.8023883,"threshold_uncertainty_score":0.6143785,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08670942664244706,"score_gpt":0.3252280140823825,"score_spread":0.2385185874399354,"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."}}