{"id":"W4282932896","doi":"10.1038/s41597-022-01463-7","title":"Datasets for benchmarking antimicrobial resistance genes in bacterial metagenomic and whole genome sequencing","year":2022,"lang":"en","type":"article","venue":"Scientific Data","topic":"Genomics and Phylogenetic Studies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":22,"is_retracted":false,"has_abstract":true,"ca_institutions":"Dalhousie University; Sunnybrook Health Science Centre; Public Health Agency of Canada; Sunnybrook Hospital; McMaster University","funders":"Biotechnology and Biological Sciences Research Council; Medical Research Council; Joint Programming Initiative on Antimicrobial Resistance; Canadian Institutes of Health Research; Quadram Institute Bioscience; Universiteit van Amsterdam; Government of Canada","keywords":"Metagenomics; Computational biology; Biology; Genome; Gene prediction; Genomics; Whole genome sequencing; Identification (biology); DNA sequencing; Gene; Genetics","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":[0.001012704,0.0001115683,0.0001331912,0.00006555617,0.0005065034,0.0001245575,0.0006818373,0.0000261766,0.00001582811],"category_scores_gemma":[0.00003507952,0.0001240627,0.00002371647,0.000106034,0.0001236509,0.000004011645,0.002092178,0.00004831922,0.000001087399],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002821559,"about_ca_system_score_gemma":0.0001275827,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002181739,"about_ca_topic_score_gemma":0.0006681614,"domain_scores_codex":[0.998574,0.00005414565,0.0002010709,0.0008216066,0.00008780124,0.0002613383],"domain_scores_gemma":[0.9989308,0.00001300082,0.00007579436,0.0009236963,0.00001937921,0.0000373174],"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":[0.00005329232,0.0000138394,0.0002001865,0.00002152933,0.00002334403,0.000001958601,0.00006484018,0.0001044324,0.9868191,0.00001096455,0.0121099,0.0005766159],"study_design_scores_gemma":[0.0005234929,0.00004298329,0.001395619,0.000004342598,0.00002232494,0.000007805426,0.0001457485,0.0001775002,0.02161449,0.00004819305,0.9758072,0.0002102731],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.93531,0.002148415,0.0001330023,0.00009348021,0.0009549828,0.0002999054,0.0610234,0.000002007254,0.0000347772],"genre_scores_gemma":[0.9313124,0.0001160082,0.006465722,0.00009501609,0.0002932519,0.0000474433,0.06112145,0.00002170726,0.0005270235],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.9652046,"threshold_uncertainty_score":0.5059131,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03591850040073827,"score_gpt":0.260202249174092,"score_spread":0.2242837487733537,"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."}}