{"id":"W2787925891","doi":"10.7717/peerj.5364","title":"Denoising the Denoisers: an independent evaluation of microbiome sequence error-correction approaches","year":2018,"lang":"en","type":"article","venue":"PeerJ","topic":"Genomics and Phylogenetic Studies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":375,"is_retracted":false,"has_abstract":true,"ca_institutions":"Dalhousie University","funders":"Natural Sciences and Engineering Research Council of Canada; Terry Fox Research Institute","keywords":"UniFrac; Spurious relationship; Pipeline (software); Amplicon; Cluster analysis; Operational taxonomic unit; Biology; Artificial intelligence; Computer science; False positive paradox; Metagenomics; Pattern recognition (psychology); Noise reduction; Computational biology; Data mining; Machine learning; 16S ribosomal RNA; Genetics; Gene; Polymerase chain reaction","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":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.04414162,0.003799107,0.002130748,0.003790651,0.002072731,0.002663172,0.003849042,0.003578248,0.001710987],"category_scores_gemma":[0.06975324,0.0009677915,0.00375901,0.002469253,0.001956746,0.003291937,0.005165026,0.003925847,0.001410495],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001322389,"about_ca_system_score_gemma":0.002752017,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004522436,"about_ca_topic_score_gemma":0.007638584,"domain_scores_codex":[0.9803232,0.00806446,0.001759071,0.003690065,0.005443736,0.0007194975],"domain_scores_gemma":[0.9620053,0.02205728,0.002459244,0.00605656,0.006609195,0.0008124779],"domain_codex":null,"domain_gemma":"methods","domain_candidate":"methods","domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.01001335,0.001648073,0.0775186,0.005125535,0.009033842,0.0005019943,0.002919757,0.1732554,0.08104389,0.01287952,0.01873216,0.6073278],"study_design_scores_gemma":[0.0007760826,0.003516985,0.02792783,0.0006582742,0.001873236,0.000854946,0.0009667849,0.8353158,0.09119909,0.01038055,0.02597695,0.0005535258],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2563948,0.006012913,0.7079183,0.0006238426,0.001186112,0.001812248,0.00293579,0.01836296,0.004753078],"genre_scores_gemma":[0.1909917,0.0009585267,0.7951531,0.0004373033,0.0001289781,0.001222854,0.005189919,0.004162048,0.001755477],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9558584,"threshold_uncertainty_score":0.233446,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1224366225942445,"score_gpt":0.3201658803552965,"score_spread":0.1977292577610521,"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."}}