{"id":"W2069716529","doi":"10.1186/1471-2105-14-s16-s6","title":"Random forests on Hadoop for genome-wide association studies of multivariate neuroimaging phenotypes","year":2013,"lang":"en","type":"article","venue":"BMC Bioinformatics","topic":"Genetic Associations and Epidemiology","field":"Biochemistry, Genetics and Molecular Biology","cited_by":48,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"National Institute on Aging; University of California, San Diego; Genentech; National Institutes of Health; Canadian Institutes of Health Research; University of California, Los Angeles; Servier; Imperial College London; Northern California Institute for Research and Education; Synarc; Engineering and Physical Sciences Research Council; Bayer HealthCare; Alzheimer's Disease Neuroimaging Initiative; Meso Scale Diagnostics; National University of Singapore; Medpace; Pfizer; Biogen; AstraZeneca; Bristol-Myers Squibb; Eli Lilly and Company; Novartis Pharmaceuticals Corporation; F. Hoffmann-La Roche; Amorfix Life Sciences; Alzheimer's Drug Discovery Foundation; Foundation for the National Institutes of Health","keywords":"Multivariate statistics; Random forest; Computer science; Artificial intelligence; Neuroimaging; Genome-wide association study; Machine learning; Genetic architecture; Genetic association; Quantitative trait locus; Population; Data mining; Computational biology; Biology; Single-nucleotide polymorphism; Genetics; Gene; Genotype","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0005089861,0.0001414088,0.0002807954,0.00005077374,0.00009593461,0.00001634982,0.0001274169,0.0001310363,0.000005685328],"category_scores_gemma":[0.005219018,0.0001187769,0.0001270255,0.00005788387,0.00003853266,0.000009329611,0.00006669037,0.0000542204,0.00001775874],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00003760727,"about_ca_system_score_gemma":0.00004703003,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001447438,"about_ca_topic_score_gemma":0.00003968719,"domain_scores_codex":[0.998919,0.0000633387,0.0005126183,0.0001321662,0.0001044317,0.0002684787],"domain_scores_gemma":[0.9982021,0.00071718,0.0004875051,0.0002160442,0.000335062,0.00004213086],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0003307265,0.0002346397,0.8794767,0.0007965834,0.0008283181,1.528654e-7,0.001791738,0.04219806,0.01023654,0.0004023681,0.05795806,0.00574616],"study_design_scores_gemma":[0.006646939,0.0009820368,0.8280545,0.00005865613,0.0001807719,0.000001604118,0.001527821,0.1391055,0.004422336,0.002755261,0.01562048,0.0006441172],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8143353,0.0004270603,0.1816377,0.0009218225,0.0003859834,0.001248037,0.00008670961,0.00002498488,0.0009323623],"genre_scores_gemma":[0.870077,0.0002080552,0.1263521,0.001603404,0.0001850085,0.0001829351,0.0002687009,0.00002846119,0.001094343],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.09690747,"threshold_uncertainty_score":0.6248028,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02956429377384484,"score_gpt":0.2919014667968587,"score_spread":0.2623371730230139,"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."}}