{"id":"W4414162033","doi":"10.1186/s13059-025-03754-9","title":"Diverse short tandem repeat sequences influence gene regulation in human populations","year":2025,"lang":"en","type":"article","venue":"Genome biology","topic":"Genetic Neurodegenerative Diseases","field":"Neuroscience","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto; SickKids Foundation; Hospital for Sick Children","funders":"Canadian Institutes of Health Research","keywords":"Human genetics; Gene; Regulation of gene expression; Phenotype; Sequence (biology); Gene expression; Tandem repeat; Regulatory sequence; Human genome","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":[0.00009994646,0.0001313065,0.0001570496,0.0002412291,0.000218689,0.00002217523,0.0002638955,0.00007700336,0.00007385132],"category_scores_gemma":[0.0002842795,0.0001287498,0.00004100327,0.0004679067,0.0002231818,0.00007937459,0.0001284586,0.00008449519,0.00001924046],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00007508579,"about_ca_system_score_gemma":0.00006252879,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00007269474,"about_ca_topic_score_gemma":0.0001472327,"domain_scores_codex":[0.9985632,0.0002672069,0.0002866337,0.0005498946,0.00007950155,0.0002536106],"domain_scores_gemma":[0.9994856,0.00005681334,0.00005295141,0.0003138891,0.00004410725,0.00004666613],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"observational","study_design_scores_codex":[0.000004552065,0.00002247422,0.1796494,0.000002512842,0.000001378709,0.000004708312,0.00004163358,0.002220832,0.8110687,0.006703076,0.00001259136,0.000268129],"study_design_scores_gemma":[0.0001218983,0.00004107729,0.6608896,0.000004507053,0.000007078417,0.000002291026,0.00001952632,0.00006865377,0.3308994,0.007568177,0.0002725924,0.0001052526],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9978205,0.00009292306,0.0003790159,0.0001620563,0.0002031335,0.0002784336,0.00003207969,0.00005122474,0.0009806617],"genre_scores_gemma":[0.998692,0.00002506644,0.0002005643,0.0005933768,0.00003587459,0.00004167194,0.00003642814,0.000006255177,0.0003687818],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.4812401,"threshold_uncertainty_score":0.5250265,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06541887658696383,"score_gpt":0.3371133803159713,"score_spread":0.2716945037290074,"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."}}