{"id":"W4412677259","doi":"10.1101/2025.07.26.666712","title":"Rare k-mers reveal centromere haplogroups underlying human diversity and cancer translocations","year":2025,"lang":"en","type":"preprint","venue":"bioRxiv (Cold Spring Harbor Laboratory)","topic":"Chromosomal and Genetic Variations","field":"Agricultural and Biological Sciences","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Moonshot Research and Development Program; Institute of Genetics; National Human Genome Research Institute; Japan Science and Technology Agency; Ministry of Education, Culture, Sports, Science and Technology; Japan Society for the Promotion of Science; National Institutes of Health; University of Tokyo; Japan Agency for Medical Research and Development","keywords":"Haplogroup; Chromosomal translocation; Centromere; Diversity (politics); Biology; Evolutionary biology; Genetics; Computational biology; Political science; Gene; Haplotype; Chromosome; Allele","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0005562254,0.0002062861,0.0003888156,0.001184898,0.0003871571,0.0004893193,0.0001911025,0.0003128338,0.001511067],"category_scores_gemma":[0.001431373,0.0001902103,0.0003297243,0.001283937,0.0003469986,0.000219063,0.0005901243,0.0003560065,0.0003345917],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001930658,"about_ca_system_score_gemma":0.0001447705,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002424409,"about_ca_topic_score_gemma":0.005108291,"domain_scores_codex":[0.9995573,0.00006734662,0.00003173737,0.0002606035,0.00004604411,0.00003703995],"domain_scores_gemma":[0.9991974,0.0002791968,0.0002119693,0.0001609813,0.00008034006,0.00007007793],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.000956649,0.00004510165,0.7468478,0.0002719744,0.0005137262,0.0005187063,0.0008844149,0.005248407,0.2000014,0.0009709757,0.0009740609,0.04276683],"study_design_scores_gemma":[0.00002442777,0.00008604374,0.9684896,0.00002772378,0.0001357639,0.001125104,0.000378538,0.01209584,0.01161283,0.001988372,0.004006706,0.00002899959],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9869854,0.0006998125,0.008528378,0.00005952189,0.000008798505,0.0000143105,0.002993555,0.000125784,0.0005844836],"genre_scores_gemma":[0.9924948,0.0001507671,0.004957528,0.00002688206,0.000006376968,0.00001102678,0.002038404,0.00002183147,0.000292327],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.002424409,"threshold_uncertainty_score":0.00505501,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03571833687313584,"score_gpt":0.2366476549407209,"score_spread":0.2009293180675851,"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."}}