{"id":"W2941864079","doi":"10.7717/peerj.6995","title":"ConFindr: rapid detection of intraspecies and cross-species contamination in bacterial whole-genome sequence data","year":2019,"lang":"en","type":"article","venue":"PeerJ","topic":"Genomics and Phylogenetic Studies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":147,"is_retracted":false,"has_abstract":true,"ca_institutions":"Canadian Food Inspection Agency","funders":"Canadian Food Inspection Agency","keywords":"Contamination; Computational biology; Whole genome sequencing; Genome; Biology; Computer science; Genetics; Data mining; 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":[0.01523469,0.002578356,0.002319354,0.004080655,0.002033412,0.004067684,0.003794452,0.001787213,0.009209187],"category_scores_gemma":[0.03353604,0.001988349,0.002201927,0.002583939,0.001504802,0.004140532,0.005838726,0.00359239,0.01084359],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001040075,"about_ca_system_score_gemma":0.003718836,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002542032,"about_ca_topic_score_gemma":0.00409643,"domain_scores_codex":[0.9886491,0.001971494,0.0008210461,0.002997202,0.004851167,0.0007098606],"domain_scores_gemma":[0.9881214,0.004975909,0.001708935,0.002252189,0.002185172,0.0007563753],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.001719457,0.0004040542,0.03425058,0.006566254,0.001173714,0.002064681,0.002664043,0.01329807,0.2477483,0.00831086,0.3898065,0.2919935],"study_design_scores_gemma":[0.0006251083,0.0007532353,0.04084193,0.00101577,0.000422966,0.002834359,0.0007166885,0.179999,0.2634379,0.01864493,0.4897161,0.0009920785],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.0658443,0.002010639,0.4424519,0.001503404,0.001113426,0.001134254,0.05531962,0.4199636,0.01065882],"genre_scores_gemma":[0.1213784,0.001170073,0.6506509,0.001501664,0.0003226018,0.002001446,0.1471555,0.07065395,0.005165502],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01523469,"threshold_uncertainty_score":0.08056974,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02802214653020833,"score_gpt":0.2611038732426098,"score_spread":0.2330817267124014,"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."}}