{"id":"W2970357378","doi":"10.1093/molbev/msz201","title":"Legacy Data Confound Genomics Studies","year":2019,"lang":"en","type":"review","venue":"Molecular Biology and Evolution","topic":"Genetic Associations and Epidemiology","field":"Biochemistry, Genetics and Molecular Biology","cited_by":31,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University; McGill University and Génome Québec Innovation Centre","funders":"","keywords":"Biology; Population stratification; Imputation (statistics); Spurious relationship; 1000 Genomes Project; Genomics; Genome-wide association study; Population; Data quality; Computational biology; Genome; Genetics; Evolutionary biology; Missing data; Computer science; Demography; Single-nucleotide polymorphism; Gene; Machine learning","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":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0006316651,0.0003882236,0.001124083,0.00009496247,0.0001296019,0.00002473253,0.0004800337,0.0009640994,0.000004638284],"category_scores_gemma":[0.0004739891,0.0003331526,0.0001858261,0.00009226418,0.0002668276,0.000006656475,0.0009091499,0.0002317865,0.00006246571],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00007138243,"about_ca_system_score_gemma":0.0003079922,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002052741,"about_ca_topic_score_gemma":0.00001918626,"domain_scores_codex":[0.997612,0.000497149,0.0005133048,0.0009516616,0.00004718595,0.0003787213],"domain_scores_gemma":[0.9983147,0.00006970293,0.0003618979,0.001100505,0.00008857075,0.00006467373],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.00004712691,0.0001237097,0.001201781,0.006700762,0.006017386,0.00001372994,0.00004551766,0.00003020907,0.002583583,0.009274921,0.01477284,0.9591885],"study_design_scores_gemma":[0.0001973545,0.000164531,0.00007362373,0.000209684,0.0007020434,0.00006443513,0.0000467116,0.00002668047,0.000004367961,0.0007458946,0.9973881,0.0003765844],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"review","genre_gemma":"review","genre_scores_codex":[0.0009294161,0.9915729,0.006023113,0.00006174749,0.0005718281,0.0004585281,0.0001706429,0.00001260141,0.0001992217],"genre_scores_gemma":[0.001212444,0.9916293,0.001080209,0.0001467169,0.0002894399,0.00004087209,0.005233817,0.00003829418,0.0003288987],"genre_candidate":"review","genre_consensus":"review","teacher_disagreement_score":0.9826152,"threshold_uncertainty_score":0.999912,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.09663825936812508,"score_gpt":0.4091853181264148,"score_spread":0.3125470587582897,"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."}}