{"id":"W6920853318","doi":"10.6084/m9.figshare.13364264","title":"Additional file 3 of Systematic evaluation of supervised machine learning for sample origin prediction using metagenomic sequencing data","year":2020,"lang":"en","type":"article","venue":"Figshare","topic":"Metabolomics and Mass Spectrometry Studies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Public Health Agency of Canada","funders":"","keywords":"Sample (material); Metagenomics; Supervised learning; Pattern recognition (psychology); Large sample","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":[0.006566299,0.001940958,0.001715669,0.004014306,0.001466383,0.002575953,0.003349751,0.001462591,0.7706314],"category_scores_gemma":[0.05968945,0.0009041827,0.002262712,0.006003206,0.0006668628,0.002350264,0.001973711,0.001576488,0.1364269],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001582596,"about_ca_system_score_gemma":0.003546824,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.008401045,"about_ca_topic_score_gemma":0.01880587,"domain_scores_codex":[0.9975272,0.0005737395,0.0003843296,0.0006476044,0.0006493567,0.0002177611],"domain_scores_gemma":[0.9371144,0.05286199,0.001707157,0.003151931,0.004444392,0.0007201598],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0008319461,0.0001419502,0.004254073,0.007135421,0.000279743,0.0001343098,0.0001545496,0.001528975,0.0005258151,0.0009828992,0.9714456,0.01258471],"study_design_scores_gemma":[0.01680743,0.0008390318,0.05205565,0.006256551,0.001364387,0.0009288893,0.001100126,0.01767443,0.005997062,0.02293898,0.8735471,0.0004904044],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"dataset","genre_gemma":"dataset","genre_scores_codex":[0.0002527226,0.0000182344,0.0008224456,0.00005829824,0.00002369534,0.0001064618,0.9970265,0.001322983,0.0003685522],"genre_scores_gemma":[0.00815447,0.00009528542,0.0147483,0.0003566393,0.00006928249,0.002555177,0.9675487,0.002703198,0.003768923],"genre_candidate":"dataset","genre_consensus":"dataset","teacher_disagreement_score":0.7706314,"threshold_uncertainty_score":0.3271664,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2503861662254359,"score_gpt":0.3229677432566421,"score_spread":0.07258157703120627,"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."}}