{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00261647,0.0005609817,0.0005962197,0.0008522479,0.0007381911,0.000817,0.00179924,0.001767049,0.01671787],"category_scores_gemma":[0.01043615,0.0005060621,0.001171906,0.001991952,0.0004058548,0.0005413136,0.0007592718,0.001711312,0.002364414],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001255565,"about_ca_system_score_gemma":0.001093679,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01376819,"about_ca_topic_score_gemma":0.0194779,"domain_scores_codex":[0.9993481,0.0003793156,0.00004726243,0.0001159116,0.0000696361,0.00003966573],"domain_scores_gemma":[0.9943715,0.003918958,0.0001964266,0.0007414271,0.0005231468,0.0002485541],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00270713,0.0006980941,0.08338663,0.0009573224,0.001098244,0.001743234,0.0004976642,0.7261853,0.003172459,0.0159461,0.1454035,0.01820435],"study_design_scores_gemma":[0.003127137,0.000579519,0.07342502,0.0004108431,0.0005829598,0.001802064,0.0007218515,0.703368,0.004892425,0.07365718,0.1371255,0.0003075887],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"dataset","genre_gemma":"dataset","genre_scores_codex":[0.3234853,0.0007109359,0.04918769,0.003234652,0.00029933,0.0009404464,0.612507,0.002164801,0.007469726],"genre_scores_gemma":[0.3679393,0.0005265974,0.05811432,0.0008237836,0.0000663074,0.003598445,0.5643355,0.0004595334,0.004136152],"genre_candidate":"dataset","genre_consensus":"dataset","teacher_disagreement_score":0.01671787,"threshold_uncertainty_score":0.05592686,"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."}}