{"id":"W3189624264","doi":"10.1186/s13059-021-02447-3","title":"Straglr: discovering and genotyping tandem repeat expansions using whole genome long-read sequences","year":2021,"lang":"en","type":"article","venue":"Genome biology","topic":"Genomics and Phylogenetic Studies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":111,"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; Genome British Columbia; Genome Canada","keywords":"Genotyping; Biology; Computational biology; Genome; Human genetics; Whole genome sequencing; Genetics; Trinucleotide repeat expansion; Human genome; Genotype; Gene","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":[0.0001453544,0.0002341578,0.0002840514,0.00004430617,0.0002849224,0.0000405506,0.0001585312,0.0001912826,0.00003157906],"category_scores_gemma":[0.00004480079,0.0002253615,0.00008956303,0.00009616202,0.000205155,0.00000217276,0.0004370734,0.00009573541,0.000005214483],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002800238,"about_ca_system_score_gemma":0.0001398646,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00005537892,"about_ca_topic_score_gemma":0.0001628901,"domain_scores_codex":[0.9984878,0.00008653566,0.0002788412,0.000656853,0.00005515908,0.000434758],"domain_scores_gemma":[0.9993128,0.0000182666,0.00009499797,0.0003794388,0.00008739194,0.0001070902],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"observational","study_design_scores_codex":[0.0000103906,0.00001662895,0.01314948,0.00001381137,0.0001261064,0.00001714357,0.0001544719,0.0003527279,0.9852024,0.00008968107,0.000003158723,0.0008639365],"study_design_scores_gemma":[0.002857703,0.001410049,0.5218878,0.00008987288,0.0004225228,0.001551238,0.00401224,0.0003146997,0.2095265,0.003249288,0.2514423,0.003235799],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9688015,0.02772833,0.002439925,0.0001306866,0.0002286378,0.0001271626,0.000127629,0.000007385726,0.0004087196],"genre_scores_gemma":[0.9939237,0.002498864,0.002380135,0.0002234781,0.000393287,0.00001166537,0.0002288826,0.00002712521,0.0003128192],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.775676,"threshold_uncertainty_score":0.9189976,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02496703257416895,"score_gpt":0.2674932000447482,"score_spread":0.2425261674705793,"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."}}