{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.02016216,0.001016611,0.002239291,0.003526454,0.0006222068,0.003212922,0.00247452,0.003844857,0.004889812],"category_scores_gemma":[0.03953711,0.0005122479,0.001263545,0.003804388,0.003590527,0.005226064,0.003090678,0.005945737,0.002682272],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001801987,"about_ca_system_score_gemma":0.003996218,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00175838,"about_ca_topic_score_gemma":0.002307496,"domain_scores_codex":[0.9942157,0.003085104,0.0005732019,0.0008468415,0.001106393,0.0001728267],"domain_scores_gemma":[0.9464213,0.04432465,0.002187198,0.002047098,0.00413237,0.0008873968],"domain_codex":null,"domain_gemma":"methods","domain_candidate":"methods","domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"systematic_review","study_design_scores_codex":[0.00008585287,0.00002075982,0.0009269508,0.01099296,0.0003838229,0.0002112662,0.0002863002,0.0002387681,0.0001656407,0.02830558,0.07293074,0.8854514],"study_design_scores_gemma":[0.00001938849,0.0000327254,0.001549101,0.01425396,0.0002052131,0.001028708,0.0001881942,0.00009064211,0.0001536119,0.02469197,0.9577442,0.00004229932],"study_design_candidate":"systematic_review","study_design_consensus":null,"genre_codex":"review","genre_gemma":"review","genre_scores_codex":[0.0001384476,0.9872795,0.001092749,0.008550948,0.001498467,0.000009876272,0.0000990499,0.0000248177,0.00130602],"genre_scores_gemma":[0.002112714,0.9821086,0.001334745,0.009610928,0.003787673,0.00003100543,0.0002037914,0.00003134233,0.0007791466],"genre_candidate":"review","genre_consensus":"review","teacher_disagreement_score":0.9798378,"threshold_uncertainty_score":0.106629,"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."}}