{"id":"W4281571943","doi":"10.1016/j.dib.2022.108311","title":"Datasets for a simulated family-based exome-sequencing study","year":2022,"lang":"en","type":"article","venue":"Data in Brief","topic":"Genetic factors in colorectal cancer","field":"Medicine","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Simon Fraser University","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Exome sequencing; Exome; Inheritance (genetic algorithm); Computational biology; Genetics; Biology; Population; Scripting language; Bioinformatics; Computer science; Medicine; Gene; Mutation","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":[0.0007240285,0.0001363652,0.0002675137,0.0001429484,0.0001094823,0.0000172143,0.000565827,0.00003238211,0.0002967226],"category_scores_gemma":[0.0004513781,0.0001418656,0.00002575756,0.0004018093,0.00003352596,0.00006841682,0.0006280734,0.0002506798,0.000006029431],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000368615,"about_ca_system_score_gemma":0.0004053446,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0008874519,"about_ca_topic_score_gemma":0.000243878,"domain_scores_codex":[0.9983394,0.00008696124,0.0003045668,0.000573481,0.0004141409,0.0002814637],"domain_scores_gemma":[0.9980235,0.0002297024,0.00006589235,0.001574,0.00003164535,0.00007527927],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"not_applicable","study_design_scores_codex":[0.01031236,0.007465113,0.4395674,0.000900049,0.0006106546,0.002265539,0.003143254,0.05784541,0.06303846,0.00004501311,0.3890235,0.02578316],"study_design_scores_gemma":[0.02766616,0.006791391,0.1508855,0.0001183005,0.0004182887,0.00005275335,0.004964693,0.1963717,0.001909631,0.0001084127,0.6097589,0.0009542429],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9822981,0.0001616653,0.0002462021,0.000175577,0.0003671641,0.00203669,0.01456286,0.00005616231,0.00009551396],"genre_scores_gemma":[0.9755371,0.000001528873,0.001558949,0.001534124,0.00005392678,0.000180375,0.0210723,0.00003266252,0.00002900885],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.2886819,"threshold_uncertainty_score":0.5785111,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1234527471407007,"score_gpt":0.3685542919375572,"score_spread":0.2451015447968565,"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."}}