{"id":"W4399850991","doi":"10.1101/2024.06.19.24309173","title":"A comprehensive tandem repeat catalog of the human genome","year":2024,"lang":"en","type":"preprint","venue":"medRxiv","topic":"Genetic Neurodegenerative Diseases","field":"Neuroscience","cited_by":11,"is_retracted":false,"has_abstract":true,"ca_institutions":"BC Children's Hospital; University of British Columbia; Canada's Michael Smith Genome Sciences Centre","funders":"Canadian Institutes of Health Research; Alliance de recherche numérique du Canada","keywords":"Tandem repeat; Genome; Human genome; Genotyping; Biology; Microsatellite; Genetics; Computational biology; STR multiplex system; Gene; Genotype; Allele","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":[0.00008542359,0.0003410518,0.0003849712,0.0001063573,0.0001496764,0.00006899536,0.001195381,0.0001182135,0.00007238871],"category_scores_gemma":[0.0002955105,0.0002420429,0.0003405726,0.0002660647,0.0004538902,0.00001362713,0.00320774,0.0006723928,0.00008996537],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00004265231,"about_ca_system_score_gemma":0.0001726946,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002618613,"about_ca_topic_score_gemma":0.00001100035,"domain_scores_codex":[0.9973275,0.0004241923,0.0004282323,0.001018705,0.0005273432,0.0002740141],"domain_scores_gemma":[0.9979483,0.0001326867,0.0002449507,0.001477738,0.0001033561,0.00009300796],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.000002619535,0.00005063551,0.003408386,0.0002594352,0.00001492546,0.00004006156,0.0003022793,0.0005533709,0.9943622,0.0007354078,0.0002460663,0.00002458466],"study_design_scores_gemma":[0.00009982051,0.00002732397,0.1035076,0.00009598151,0.00009100178,0.00001166074,0.00001795599,0.00004832661,0.8873349,0.006986687,0.001544501,0.0002342165],"study_design_candidate":"bench_or_experimental","study_design_consensus":"bench_or_experimental","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9945847,0.0007438785,0.000007246913,0.000392355,0.001349642,0.0006609234,0.0003599209,0.00008792755,0.0018134],"genre_scores_gemma":[0.9978676,0.00004943264,0.00001960417,0.0005174102,0.0001522851,0.00007612506,0.0000131462,0.0000577981,0.001246601],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.1070273,"threshold_uncertainty_score":0.9870223,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06007383145162026,"score_gpt":0.3001326257769286,"score_spread":0.2400587943253084,"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."}}