{"id":"W4293222524","doi":"10.1101/2022.08.24.505159","title":"A genome-wide atlas of recurrent repeat expansions in human cancer","year":2022,"lang":"en","type":"preprint","venue":"bioRxiv (Cold Spring Harbor Laboratory)","topic":"Genetic Neurodegenerative Diseases","field":"Neuroscience","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"Hospital for Sick Children; University of Toronto","funders":"Stanford Cancer Institute; National Institutes of Health","keywords":"Biology; Human genome; Genome; Genetics; Tandem repeat; Gene; Microsatellite; Cancer; Computational biology; DNA sequencing; Trinucleotide repeat expansion; 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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0004214139,0.0002819905,0.0004774568,0.002635635,0.0003258174,0.0004737461,0.0002994071,0.0003043059,0.002464356],"category_scores_gemma":[0.0006033596,0.0001531782,0.0003352175,0.00332962,0.0001279756,0.0001271513,0.0005462049,0.000299181,0.0007587926],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003350686,"about_ca_system_score_gemma":0.000479036,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002334774,"about_ca_topic_score_gemma":0.005436224,"domain_scores_codex":[0.999602,0.00005951093,0.0000302956,0.0001750188,0.00008910573,0.00004409855],"domain_scores_gemma":[0.9996095,0.00009012461,0.00009547403,0.00006982878,0.00006169889,0.00007340927],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.00165736,0.000120498,0.4279902,0.003613085,0.002979952,0.002159416,0.0004460453,0.004120606,0.3118367,0.00372633,0.09573769,0.1456121],"study_design_scores_gemma":[0.00007630011,0.0001245141,0.8894225,0.00008872907,0.0005761833,0.003041802,0.0001098253,0.002341722,0.0140965,0.001437223,0.08865286,0.00003189249],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.644592,0.02048313,0.008970614,0.0007840637,0.000119509,0.0001271119,0.3132478,0.001888985,0.009786782],"genre_scores_gemma":[0.6577535,0.00725286,0.01552371,0.0004229765,0.0001153153,0.0002832211,0.3145207,0.0002800715,0.003847693],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.002635635,"threshold_uncertainty_score":0.008244097,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03405718543823828,"score_gpt":0.2738546715626531,"score_spread":0.2397974861244148,"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."}}