{"id":"W4313268822","doi":"10.1101/2022.12.15.520583","title":"A schema for digitized surface swab site metadata in open-source DNA sequence databases","year":2022,"lang":"en","type":"preprint","venue":"bioRxiv (Cold Spring Harbor Laboratory)","topic":"Data-Driven Disease Surveillance","field":"Medicine","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Simon Fraser University","funders":"National Institute of Food and Agriculture; U.S. Department of Agriculture","keywords":"Metadata; Computer science; Information retrieval; Controlled vocabulary; 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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.01298299,0.0006562339,0.0008316966,0.01264072,0.001696384,0.00790517,0.001986358,0.00179481,0.003597108],"category_scores_gemma":[0.02116718,0.0009549896,0.001644249,0.01286024,0.001762908,0.009627122,0.003754294,0.001741364,0.003103958],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.003580855,"about_ca_system_score_gemma":0.006710998,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01062607,"about_ca_topic_score_gemma":0.008807686,"domain_scores_codex":[0.9926563,0.001460001,0.003153032,0.0009851045,0.001486131,0.0002593676],"domain_scores_gemma":[0.9782085,0.007595171,0.003333002,0.004494601,0.005682473,0.0006862084],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"not_applicable","study_design_scores_codex":[0.000369238,0.0005558705,0.03836496,0.002961742,0.0001958096,0.002434785,0.01331602,0.01655792,0.02933986,0.5797162,0.05651603,0.2596715],"study_design_scores_gemma":[0.0001037959,0.0001897934,0.01175096,0.002318113,0.0001425895,0.001544627,0.006043589,0.04942455,0.01696028,0.1006715,0.8106328,0.0002175952],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.02490504,0.0004779048,0.8916181,0.001735936,0.0001919763,0.002990077,0.05877541,0.009007595,0.01029805],"genre_scores_gemma":[0.05109297,0.0006253535,0.853477,0.0004629687,0.00004947857,0.001456707,0.08959857,0.000981995,0.00225492],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01298299,"threshold_uncertainty_score":0.06866145,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05559735390608114,"score_gpt":0.3067777213465103,"score_spread":0.2511803674404292,"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."}}