{"id":"W4398141994","doi":"10.7554/elife.98300","title":"Benchmarking reveals superiority of deep learning variant callers on bacterial nanopore sequence data","year":2024,"lang":"en","type":"article","venue":"eLife","topic":"Genomics and Phylogenetic Studies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":22,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Institute of Infection and Immunity; University of Pittsburgh; National Health and Medical Research Council; University of Melbourne; Medical Research Council; Australian Government","keywords":"Nanopore sequencing; Benchmarking; Deep sequencing; DNA sequencing; Computational biology; Precision and recall; Genomics; Computer science; Deep learning; Genome; Artificial intelligence; Metagenomics; Ion semiconductor sequencing; Identification (biology); Biology; Data mining; Machine learning; Genetics; Gene","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0002686311,0.0001160227,0.000141162,0.00002296306,0.00007075688,0.00002676187,0.0002459282,0.0000813415,0.00003534135],"category_scores_gemma":[0.0001072497,0.0001058972,0.00004545684,0.00005212015,0.00005789067,0.000001152485,0.000301171,0.00009006747,0.000004741604],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000009553206,"about_ca_system_score_gemma":0.00004857885,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00003498327,"about_ca_topic_score_gemma":0.00003261258,"domain_scores_codex":[0.9990972,0.00005101652,0.0001865602,0.0003826875,0.0001157678,0.0001667528],"domain_scores_gemma":[0.9994666,0.0000237116,0.00003911864,0.0003866445,0.00004040786,0.00004353496],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"not_applicable","study_design_scores_codex":[0.00003313148,0.00001307783,0.002675693,0.00004090046,0.0001002223,0.00001508738,0.0001495583,0.0001441261,0.9916709,0.0001161215,0.0004738588,0.00456734],"study_design_scores_gemma":[0.0009016935,0.001695701,0.03867473,0.000296958,0.0001614818,0.00009055731,0.0004134689,0.005966279,0.3063277,0.0001352052,0.6442776,0.00105866],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9959512,0.002206468,0.0002455261,0.00007767048,0.0006233635,0.00008449646,0.0001090798,0.000005661949,0.0006965605],"genre_scores_gemma":[0.9966242,0.001127736,0.001227885,0.0000643267,0.0005840886,0.00000405927,0.000254949,0.00001574685,0.00009699596],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.6853432,"threshold_uncertainty_score":0.4318362,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03393038422169525,"score_gpt":0.2842638463832477,"score_spread":0.2503334621615524,"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."}}