{"id":"W6901960049","doi":"10.6084/m9.figshare.13364267","title":"Additional file 4 of Systematic evaluation of supervised machine learning for sample origin prediction using metagenomic sequencing data","year":2020,"lang":"en","type":"article","venue":"Figshare","topic":"Species Distribution and Climate Change","field":"Environmental Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Public Health Agency of Canada","funders":"","keywords":"Sample (material); Multivariate statistics; Regression; Geographic coordinate system; Metagenomics; Longitude; Pattern recognition (psychology); Line (geometry); Data point","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":["insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.00876577,0.001927401,0.001849236,0.003522953,0.00117507,0.002273524,0.00400523,0.001680375,0.7841244],"category_scores_gemma":[0.08052,0.0009710311,0.002447336,0.005085827,0.0006261888,0.002529463,0.001819576,0.001523554,0.1630215],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001235788,"about_ca_system_score_gemma":0.00312358,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006701971,"about_ca_topic_score_gemma":0.01785132,"domain_scores_codex":[0.9971903,0.0008492517,0.0003964621,0.000724618,0.0006235619,0.0002157833],"domain_scores_gemma":[0.9114935,0.07627077,0.002118587,0.004235771,0.005075713,0.0008055382],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0005912814,0.000114776,0.003962595,0.006373703,0.0002705588,0.0000938282,0.0001023096,0.00174694,0.0002736989,0.0007823705,0.9724821,0.01320583],"study_design_scores_gemma":[0.01513136,0.000923514,0.05209394,0.008028693,0.00123937,0.0008756735,0.0008116344,0.02679921,0.004206917,0.02225091,0.8671258,0.0005128794],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"dataset","genre_gemma":"dataset","genre_scores_codex":[0.0003451795,0.00003287514,0.001455894,0.00009055734,0.00004641844,0.0001382447,0.9954634,0.001979202,0.0004481099],"genre_scores_gemma":[0.01306128,0.000140404,0.02569837,0.0006196884,0.0001372784,0.003657377,0.9461313,0.004336854,0.00621755],"genre_candidate":"dataset","genre_consensus":"dataset","teacher_disagreement_score":0.7841244,"threshold_uncertainty_score":0.3079203,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.4158438767166555,"score_gpt":0.3267800331418235,"score_spread":0.08906384357483205,"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."}}