{"id":"W4385392633","doi":"10.1101/2023.07.27.550528","title":"An Optimized Pipeline for Detection of Salmonella Sequences in Shotgun Metagenomics Datasets","year":2023,"lang":"en","type":"preprint","venue":"bioRxiv (Cold Spring Harbor Laboratory)","topic":"Genomics and Phylogenetic Studies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"Canadian Food Inspection Agency; Carleton University","funders":"Alliance de recherche numérique du Canada; Ontario Ministry of Agriculture, Food and Rural Affairs; Canadian Food Inspection Agency","keywords":"Metagenomics; False positive paradox; Shotgun sequencing; Pipeline (software); Shotgun; Sensitivity (control systems); Computational biology; Salmonella; Biology; Computer science; Data mining; DNA sequencing; Artificial intelligence; Genetics; Bacteria; Engineering; 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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.003320422,0.002724172,0.00122348,0.002113679,0.001242016,0.001986241,0.001761519,0.001602434,0.009218128],"category_scores_gemma":[0.006141726,0.00164581,0.002427554,0.0014794,0.0006026863,0.001481912,0.001665615,0.002377079,0.007765443],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001027043,"about_ca_system_score_gemma":0.001985631,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002797348,"about_ca_topic_score_gemma":0.003997604,"domain_scores_codex":[0.9986132,0.0002356765,0.0001464707,0.0005821215,0.0002782374,0.0001442788],"domain_scores_gemma":[0.9985532,0.0005521509,0.0001195892,0.0002828137,0.0003676123,0.0001246494],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.004162128,0.0007456489,0.01658334,0.003693922,0.001589585,0.00135106,0.001120641,0.02686557,0.5213359,0.004888649,0.09976099,0.3179026],"study_design_scores_gemma":[0.0009061896,0.0006249203,0.02508548,0.0002075938,0.000422711,0.00105416,0.0003369246,0.5735396,0.3159153,0.01706746,0.06442375,0.0004158667],"study_design_candidate":"bench_or_experimental","study_design_consensus":"bench_or_experimental","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.03303815,0.0004086085,0.7111195,0.0004053102,0.0002225651,0.0009365911,0.02750247,0.225048,0.001318698],"genre_scores_gemma":[0.07613623,0.000228551,0.8628284,0.000406383,0.00006424849,0.001448964,0.04777734,0.009263825,0.001846052],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.009218128,"threshold_uncertainty_score":0.03083771,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02438678104016524,"score_gpt":0.2560042114812648,"score_spread":0.2316174304410996,"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."}}