{"id":"W4302016097","doi":"10.1101/2022.10.03.508010","title":"The variant catalogue pipeline: A workflow to generate a background variant library from Whole Genome Sequences","year":2022,"lang":"en","type":"preprint","venue":"bioRxiv (Cold Spring Harbor Laboratory)","topic":"Genomics and Rare Diseases","field":"Biochemistry, Genetics and Molecular Biology","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"BC Children's Hospital; University of British Columbia","funders":"Provincial Health Services Authority; Genome British Columbia; Michael Smith Health Research BC; BC Children's Hospital; Children's Hospital Foundation; Canadian Institutes of Health Research; Genome Canada","keywords":"Pipeline (software); Workflow; Annotation; Computational biology; Genome; 1000 Genomes Project; Biology; Computer science; Genetics; Single-nucleotide polymorphism; Database; Gene; Genotype; Programming language","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":[0.005909435,0.002805188,0.001923652,0.006211283,0.001627803,0.00414814,0.003131415,0.00193491,0.04817024],"category_scores_gemma":[0.01287255,0.002630438,0.00310935,0.003131518,0.0006834512,0.001940333,0.003675964,0.003413917,0.04243838],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001024333,"about_ca_system_score_gemma":0.004132994,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005609446,"about_ca_topic_score_gemma":0.005852662,"domain_scores_codex":[0.9979919,0.0002461676,0.0002733982,0.000741608,0.0005711966,0.0001757679],"domain_scores_gemma":[0.9963565,0.001563458,0.0002939554,0.0007974138,0.0006601696,0.000328532],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.002514479,0.0002638854,0.009497819,0.003015428,0.0009852453,0.001656311,0.0008108175,0.008982288,0.05962821,0.01308537,0.6161694,0.2833906],"study_design_scores_gemma":[0.001879077,0.0003341266,0.01657199,0.0007281741,0.0004965181,0.002651341,0.0003096756,0.08592046,0.09490472,0.06748579,0.7278636,0.0008545034],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.004535032,0.0006507413,0.4898887,0.0006708732,0.0004489323,0.000903202,0.1583134,0.339083,0.005506014],"genre_scores_gemma":[0.02296536,0.0007261678,0.5730546,0.001055124,0.0002190834,0.002296666,0.3326605,0.05975521,0.007267369],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.04817024,"threshold_uncertainty_score":0.1611455,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01150030529608134,"score_gpt":0.2101035433986109,"score_spread":0.1986032381025296,"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."}}