{"id":"W4280491254","doi":"10.1002/sim.9431","title":"The impact of methodological choices when developing predictive models using urinary metabolite data","year":2022,"lang":"en","type":"article","venue":"Statistics in Medicine","topic":"Metabolomics and Mass Spectrometry Studies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta; University of British Columbia","funders":"Natural Sciences and Engineering Research Council of Canada; Canadian Institutes of Health Research; Canada Foundation for Innovation; National Institute of Diabetes and Digestive and Kidney Diseases; Kidney Foundation of Canada","keywords":"Metabolomics; Computer science; Generalizability theory; Database normalization; Normalization (sociology); Linear discriminant analysis; Metabolite; Urinary system; Data mining; Computational biology; Bioinformatics; Medicine; Artificial intelligence; Biology; Internal medicine; Pattern recognition (psychology); Statistics; Mathematics","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":["metaresearch"],"consensus_categories":["metaresearch"],"category_scores_codex":[0.2475854,0.001692635,0.001663018,0.002179062,0.001317487,0.00613237,0.004000113,0.00203726,0.001161],"category_scores_gemma":[0.4148795,0.0006658991,0.003376356,0.003017907,0.002023696,0.003528899,0.002747273,0.004208668,0.0005322772],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001350752,"about_ca_system_score_gemma":0.003760666,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004669586,"about_ca_topic_score_gemma":0.003468517,"domain_scores_codex":[0.8586358,0.1221062,0.007985769,0.005284259,0.005335683,0.0006522037],"domain_scores_gemma":[0.6046245,0.358834,0.007889806,0.01781592,0.01018127,0.0006544138],"domain_codex":null,"domain_gemma":"methods","domain_candidate":"methods","domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.003731789,0.0007570719,0.1318163,0.00255845,0.01015964,0.001079796,0.00089727,0.4334115,0.006289486,0.04343954,0.006819921,0.3590393],"study_design_scores_gemma":[0.0008543691,0.00100369,0.02694663,0.001185603,0.001393403,0.0005939154,0.0006945092,0.8491929,0.01017037,0.09820911,0.00950698,0.0002485026],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.06234834,0.00352497,0.9223659,0.00597685,0.0005246981,0.001122718,0.001631988,0.0006880251,0.001816569],"genre_scores_gemma":[0.531747,0.002061336,0.4573217,0.002252878,0.0004238096,0.002327843,0.002834278,0.000431913,0.0005992411],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.7524146,"threshold_uncertainty_score":0.9278612,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1821100427136131,"score_gpt":0.4313333413105701,"score_spread":0.249223298596957,"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."}}