{"id":"W4414789830","doi":"10.1101/2025.10.02.25336626","title":"Whole-genome sequencing of 3,135 individuals representing the genetic diversity of the Japanese population","year":2025,"lang":"en","type":"preprint","venue":"medRxiv","topic":"Forensic and Genetic Research","field":"Biochemistry, Genetics and Molecular Biology","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University; McGill University and Génome Québec Innovation Centre; McGill Genome Centre","funders":"Japan Society for the Promotion of Science; Ministry of Health, Labour and Welfare; Japan Agency for Medical Research and Development","keywords":"Imputation (statistics); Genetic diversity; Human genetic variation; Identity by descent; Minor allele frequency; Haplotype; Genetic variation; Genetic variants; Allele frequency; Allele","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":[],"consensus_categories":[],"category_scores_codex":[0.0007421299,0.0001685721,0.0002508161,0.00005746277,0.0001882869,0.00001507623,0.001043918,0.0002547296,0.00001210005],"category_scores_gemma":[0.0003547744,0.0001096087,0.0002474674,0.0001654617,0.0002745148,0.000001381219,0.004945178,0.000290734,8.684285e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002281906,"about_ca_system_score_gemma":0.0001685339,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001074678,"about_ca_topic_score_gemma":0.0000829155,"domain_scores_codex":[0.9982277,0.0003087816,0.0003923114,0.0004101286,0.0004467188,0.0002143606],"domain_scores_gemma":[0.9981072,0.00004279169,0.0003565595,0.001253124,0.0002065004,0.00003380612],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.00001873569,0.00002229718,0.8246278,0.0003155519,0.0002358881,9.53437e-7,0.001008625,0.006329268,0.1666355,0.00002076398,0.0002087273,0.0005759873],"study_design_scores_gemma":[0.0001838434,0.0000363099,0.9503632,0.0001202694,0.000103545,0.000003019095,0.0003402345,0.0001712243,0.0472742,0.0006204777,0.0006371558,0.0001464802],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.996986,0.001180708,0.00007628082,0.0002853863,0.0002165977,0.0004483772,0.0001220669,0.000003905495,0.0006806527],"genre_scores_gemma":[0.9978145,0.0001234005,0.0002013625,0.00005173859,0.0001095627,0.00001237097,0.0001136065,0.000009837709,0.001563671],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.1257355,"threshold_uncertainty_score":0.616381,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0290440031717261,"score_gpt":0.2834822904750134,"score_spread":0.2544382873032873,"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."}}