{"id":"W1970546246","doi":"10.1534/genetics.114.162149","title":"Predicting Discovery Rates of Genomic Features","year":2014,"lang":"en","type":"article","venue":"Genetics","topic":"Genetic Associations and Epidemiology","field":"Biochemistry, Genetics and Molecular Biology","cited_by":18,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University and Génome Québec Innovation Centre","funders":"National Heart, Lung, and Blood Institute; Canada Research Chairs","keywords":"Jackknife resampling; Biology; Population; 1000 Genomes Project; Sample size determination; Exome sequencing; Exome; Selection (genetic algorithm); Computational biology; Estimator; Genome; Sample (material); Genetics; Computer science; Statistics; Single-nucleotide polymorphism; Mathematics; Machine learning; Mutation","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0002411648,0.00009855815,0.0001483223,0.00002449056,0.00004911184,0.000009667268,0.0001496173,0.0001346434,0.000006395652],"category_scores_gemma":[0.0001851564,0.00009259496,0.00007103735,0.00003532941,0.00006316412,0.000001107581,0.0001024995,0.00005170335,0.000005326473],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000005747923,"about_ca_system_score_gemma":0.00003418723,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001495176,"about_ca_topic_score_gemma":0.00006021154,"domain_scores_codex":[0.9992365,0.00008844086,0.0002194018,0.0002018817,0.00006116922,0.0001925534],"domain_scores_gemma":[0.9994594,0.00003407576,0.000131589,0.0002797329,0.00005603943,0.00003920382],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.00001066406,0.00002152888,0.5734411,0.00001182035,0.0000474603,8.817079e-8,0.000036341,0.001994139,0.4198229,0.0001073785,0.002359892,0.002146686],"study_design_scores_gemma":[0.0002891929,0.000260015,0.9011335,0.00000521368,0.00002640748,0.000005996824,0.00006708906,0.000546289,0.08726224,0.0005054814,0.00975208,0.0001464434],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.992441,0.0009046928,0.005119415,0.00009990597,0.0001471883,0.00007326762,0.00001746615,0.000005204887,0.001191856],"genre_scores_gemma":[0.9932234,0.0002531884,0.005325374,0.0001719656,0.0003086138,0.000005987593,0.00006590918,0.00001528912,0.0006302728],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.3325606,"threshold_uncertainty_score":0.3775913,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0091977577857734,"score_gpt":0.2564356339807236,"score_spread":0.2472378761949502,"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."}}