{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,3]],"date-time":"2026-09-03T18:14:37Z","timestamp":1788459277364,"version":"build-2803163510"},"reference-count":40,"publisher":"Public Library of Science (PLoS)","issue":"9","license":[{"start":{"date-parts":[[2026,9,3]],"date-time":"2026-09-03T00:00:00Z","timestamp":1788393600000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["www.plosone.org"],"crossmark-restriction":false},"short-container-title":["PLoS One"],"abstract":"<jats:sec id=\"sec001\">\n                    <jats:title>Background<\/jats:title>\n                    <jats:p>High AUROC is often taken as evidence that a transcriptomic classifier is promising, but rank discrimination can conceal fixed-threshold failure after cohort or platform transfer.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec id=\"sec002\">\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>We benchmarked four GEO whole-blood cohorts: GSE65682 for discovery, GSE95233 for external microarray validation, GSE154918 for cross-platform RNA-seq validation, and GSE28750 for non-infectious inflammation stress testing. We compared logistic-regression workflows using training-derived standard scaling, training-derived robust scaling, sample-wise rank normalization with training-derived scaling, and robust scaling using unsupervised external-cohort reference statistics. Internal performance used five-fold cross-validation with fold-contained imputation and scaling. External uncertainty used 2,000 stratified bootstrap replicates.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec id=\"sec003\">\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>Internal discrimination was very high for all strategies, but external validation revealed threshold collapse for training-derived standard and robust scaling. In GSE154918, both had balanced accuracy 0.50 at the 0.5 threshold despite very high AUROC, equivalent to random classification at that fixed threshold. The strict-inductive sample-rank strategy preserved fixed-threshold performance across external cohorts (balanced accuracy 0.95\u20131.00). Robust external-cohort adaptation also performed well (0.95\u20131.00) but uses unlabeled external-cohort distribution statistics and is therefore reported as adaptation rather than fixed single-sample transfer. Calibration and regularization sensitivity did not rescue the failing training-derived scaling strategies. In the sepsis-versus-non-infectious-inflammation stress test, robust external-cohort adaptation had the highest observed balanced accuracy (0.80, 95% CI 0.61\u20130.95), but its difference from sample-rank normalization was uncertain in paired bootstrap analysis.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec id=\"sec004\">\n                    <jats:title>Conclusions<\/jats:title>\n                    <jats:p>High AUROC can mask fixed-threshold failure in sepsis transcriptomic classifiers. In this benchmark, strict-inductive sample-rank normalization was the most stable fixed external strategy, while robust external-cohort scaling was best interpreted as unsupervised cohort adaptation whose reliability depends on external reference-sample availability.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.1371\/journal.pone.0357585","type":"journal-article","created":{"date-parts":[[2026,9,3]],"date-time":"2026-09-03T17:29:16Z","timestamp":1788456556000},"page":"e0357585","update-policy":"https:\/\/doi.org\/10.1371\/journal.pone.corrections_policy","source":"Crossref","is-referenced-by-count":0,"title":["High AUROC can mask decision failure in sepsis transcriptomic classifiers: Preprocessing stability outweighs post hoc calibration across cohorts"],"prefix":"10.1371","volume":"21","author":[{"ORCID":"https:\/\/orcid.org\/0009-0002-6221-9828","authenticated-orcid":true,"given":"Hongwei","family":"Zheng","sequence":"first","affiliation":[{"name":"Quanzhou Medical College","place":["Quanzhou, Fujian, People\u2019s Republic of China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Wenbiao","family":"Chen","sequence":"additional","affiliation":[{"name":"Quanzhou Medical College","place":["Quanzhou, Fujian, People\u2019s Republic of China"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"340","published-online":{"date-parts":[[2026,9,3]]},"reference":[{"key":"pone.0357585.ref001","article-title":"The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3)","author":"M Singer","year":"2016","journal-title":"JAMA"},{"key":"pone.0357585.ref002","article-title":"Global, regional, and national sepsis incidence and mortality, 1990-2017: analysis for the Global Burden of Disease Study","author":"KE Rudd","year":"2020","journal-title":"Lancet"},{"issue":"1","key":"pone.0357585.ref003","doi-asserted-by":"crossref","first-page":"238","DOI":"10.1186\/s13054-024-05032-9","article-title":"Enhancing sepsis biomarker development: key considerations from public and private perspectives","volume":"28","author":"J-F Llitjos","year":"2024","journal-title":"Crit Care"},{"issue":"3","key":"pone.0357585.ref004","doi-asserted-by":"crossref","DOI":"10.1186\/cc10274","article-title":"Development and validation of a novel molecular biomarker diagnostic test for the early detection of sepsis","volume":"15","author":"A Sutherland","year":"2011","journal-title":"Crit Care"},{"key":"pone.0357585.ref005","article-title":"Integrative multi-omics and machine learning identify mitochondrial biomarkers for pathogen-specific sepsis stratification and translational prioritization","author":"C Jin","year":"2026","journal-title":"Eur J Med Res"},{"key":"pone.0357585.ref006","doi-asserted-by":"crossref","first-page":"1630446","DOI":"10.3389\/fcimb.2025.1630446","article-title":"Machine learning-based identification of leptin-associated biomarkers and prognostic prediction models in sepsis","volume":"15","author":"X Liu","year":"2025","journal-title":"Front Cell Infect Microbiol"},{"issue":"5","key":"pone.0357585.ref007","doi-asserted-by":"crossref","first-page":"325","DOI":"10.1016\/S2352-4642(24)00017-8","article-title":"Host gene expression signatures to identify infection type and organ dysfunction in children evaluated for sepsis: a multicentre cohort study","volume":"8","author":"LJ Schlapbach","year":"2024","journal-title":"Lancet Child Adolesc Health"},{"issue":"12","key":"pone.0357585.ref008","doi-asserted-by":"crossref","first-page":"4119","DOI":"10.1038\/s41591-025-03964-5","article-title":"A consensus blood transcriptomic framework for sepsis","volume":"31","author":"BP Scicluna","year":"2025","journal-title":"Nat Med"},{"key":"pone.0357585.ref009","doi-asserted-by":"crossref","DOI":"10.1136\/bmj.g7594","article-title":"Transparent reporting of a multivariable prediction model for individual prognosis or diagnosis (TRIPOD): the TRIPOD statement","volume":"350","author":"GS Collins","year":"2015","journal-title":"BMJ"},{"key":"pone.0357585.ref010","doi-asserted-by":"crossref","DOI":"10.1136\/bmj-2023-078378","article-title":"TRIPOD+AI statement: updated guidance for reporting clinical prediction models that use regression or machine learning methods","volume":"385","author":"GS Collins","year":"2024","journal-title":"BMJ"},{"issue":"1","key":"pone.0357585.ref011","doi-asserted-by":"crossref","first-page":"230","DOI":"10.1186\/s12916-019-1466-7","article-title":"Calibration: the Achilles heel of predictive analytics","volume":"17","author":"Topic Group \u2018Evaluating diagnostic tests and prediction models\u2019 of the STRATOS initiative","year":"2019","journal-title":"BMC Med"},{"issue":"1","key":"pone.0357585.ref012","doi-asserted-by":"crossref","first-page":"128","DOI":"10.1097\/EDE.0b013e3181c30fb2","article-title":"Assessing the performance of prediction models: a framework for traditional and novel measures","volume":"21","author":"EW Steyerberg","year":"2010","journal-title":"Epidemiology"},{"issue":"6","key":"pone.0357585.ref013","doi-asserted-by":"crossref","first-page":"565","DOI":"10.1177\/0272989X06295361","article-title":"Decision curve analysis: a novel method for evaluating prediction models","volume":"26","author":"AJ Vickers","year":"2006","journal-title":"Med Decis Making"},{"issue":"1","key":"pone.0357585.ref014","doi-asserted-by":"crossref","first-page":"51","DOI":"10.7326\/M18-1376","article-title":"PROBAST: A Tool to Assess the Risk of Bias and Applicability of Prediction Model Studies","volume":"170","author":"RF Wolff","year":"2019","journal-title":"Ann Intern Med"},{"issue":"1","key":"pone.0357585.ref015","doi-asserted-by":"crossref","first-page":"118","DOI":"10.1093\/biostatistics\/kxj037","article-title":"Adjusting batch effects in microarray expression data using empirical Bayes methods","volume":"8","author":"WE Johnson","year":"2007","journal-title":"Biostatistics"},{"issue":"10","key":"pone.0357585.ref016","doi-asserted-by":"crossref","first-page":"733","DOI":"10.1038\/nrg2825","article-title":"Tackling the widespread and critical impact of batch effects in high-throughput data","volume":"11","author":"JT Leek","year":"2010","journal-title":"Nat Rev Genet"},{"issue":"3","key":"pone.0357585.ref017","doi-asserted-by":"crossref","first-page":"397","DOI":"10.1093\/bioinformatics\/btw650","article-title":"Improving cross-study prediction through addon batch effect adjustment or addon normalization","volume":"33","author":"R Hornung","year":"2017","journal-title":"Bioinformatics"},{"issue":"9","key":"pone.0357585.ref018","doi-asserted-by":"crossref","first-page":"1151","DOI":"10.1038\/nbt1239","article-title":"The MicroArray Quality Control (MAQC) project shows inter- and intraplatform reproducibility of gene expression measurements","volume":"24","author":"MAQC Consortium","year":"2006","journal-title":"Nat Biotechnol"},{"issue":"2","key":"pone.0357585.ref019","doi-asserted-by":"crossref","DOI":"10.1371\/journal.pone.0017238","article-title":"Removing batch effects in analysis of expression microarray data: an evaluation of six batch adjustment methods","volume":"6","author":"C Chen","year":"2011","journal-title":"PLoS One"},{"issue":"1","key":"pone.0357585.ref020","doi-asserted-by":"crossref","first-page":"2726","DOI":"10.1038\/s41598-022-06484-1","article-title":"Evaluation of domain generalization and adaptation on improving model robustness to temporal dataset shift in clinical medicine","volume":"12","author":"LL Guo","year":"2022","journal-title":"Sci Rep"},{"issue":"9","key":"pone.0357585.ref021","doi-asserted-by":"crossref","first-page":"1525","DOI":"10.1093\/jamia\/ocac093","article-title":"The harm of class imbalance corrections for risk prediction models: illustration and simulation using logistic regression","volume":"29","author":"R van den Goorbergh","year":"2022","journal-title":"J Am Med Inform Assoc"},{"issue":"1","key":"pone.0357585.ref022","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1186\/s41512-025-00216-5","article-title":"Prediction models developed using artificial intelligence: similar predictive performances with highly varying predictions for individuals - an illustration in deep vein thrombosis","volume":"10","author":"M Yusufujiang","year":"2026","journal-title":"Diagn Progn Res"},{"issue":"1","key":"pone.0357585.ref023","doi-asserted-by":"crossref","first-page":"170","DOI":"10.1186\/s12874-025-02592-4","article-title":"Extended sample size calculations for evaluation of prediction models using a threshold for classification","volume":"25","author":"R Whittle","year":"2025","journal-title":"BMC Med Res Methodol"},{"issue":"7","key":"pone.0357585.ref024","doi-asserted-by":"crossref","first-page":"407","DOI":"10.1038\/nri.2017.36","article-title":"The immunopathology of sepsis and potential therapeutic targets","volume":"17","author":"T van der Poll","year":"2017","journal-title":"Nat Rev Immunol"},{"key":"pone.0357585.ref025","article-title":"Surviving Sepsis Campaign: International Guidelines for Management of Sepsis and Septic Shock 2021","author":"L Evans","year":"2021","journal-title":"Critical Care Medicine"},{"issue":"1","key":"pone.0357585.ref026","doi-asserted-by":"crossref","first-page":"287","DOI":"10.1186\/s13054-020-02993-5","article-title":"Biomarkers of sepsis: time for a reappraisal","volume":"24","author":"C Pierrakos","year":"2020","journal-title":"Crit Care"},{"issue":"4","key":"pone.0357585.ref027","doi-asserted-by":"crossref","first-page":"259","DOI":"10.1016\/S2213-2600(16)00046-1","article-title":"Genomic landscape of the individual host response and outcomes in sepsis: a prospective cohort study","volume":"4","author":"EE Davenport","year":"2016","journal-title":"Lancet Respir Med"},{"issue":"1","key":"pone.0357585.ref028","doi-asserted-by":"crossref","first-page":"694","DOI":"10.1038\/s41467-018-03078-2","article-title":"A community approach to mortality prediction in sepsis via gene expression analysis","volume":"9","author":"TE Sweeney","year":"2018","journal-title":"Nat Commun"},{"issue":"6","key":"pone.0357585.ref029","doi-asserted-by":"crossref","first-page":"915","DOI":"10.1097\/CCM.0000000000003084","article-title":"Unsupervised Analysis of Transcriptomics in Bacterial Sepsis Across Multiple Datasets Reveals Three Robust Clusters","volume":"46","author":"TE Sweeney","year":"2018","journal-title":"Crit Care Med"},{"key":"pone.0357585.ref030","article-title":"Benchmarking sepsis gene expression diagnostics using public data","author":"TE Sweeney","year":"2017","journal-title":"Critical Care Medicine"},{"issue":"12","key":"pone.0357585.ref031","doi-asserted-by":"crossref","DOI":"10.1371\/journal.pmed.1001916","article-title":"A Molecular Host Response Assay to Discriminate Between Sepsis and Infection-Negative Systemic Inflammation in Critically Ill Patients: Discovery and Validation in Independent Cohorts","volume":"12","author":"L McHugh","year":"2015","journal-title":"PLoS Med"},{"key":"pone.0357585.ref032","doi-asserted-by":"crossref","first-page":"92","DOI":"10.1016\/j.jcrc.2018.10.028","article-title":"Validation of diagnostic gene sets to identify critically ill patients with sepsis","volume":"49","author":"DM Maslove","year":"2019","journal-title":"J Crit Care"},{"issue":"346","key":"pone.0357585.ref033","doi-asserted-by":"crossref","DOI":"10.1126\/scitranslmed.aaf7165","article-title":"Robust classification of bacterial and viral infections via integrated host gene expression diagnostics","volume":"8","author":"TE Sweeney","year":"2016","journal-title":"Sci Transl Med"},{"issue":"203","key":"pone.0357585.ref034","article-title":"A host-based RT-PCR gene expression signature to identify acute respiratory viral infection","volume":"5","author":"AK Zaas","year":"2013","journal-title":"Sci Transl Med"},{"issue":"1","key":"pone.0357585.ref035","doi-asserted-by":"crossref","first-page":"14","DOI":"10.1093\/jnci\/95.1.14","article-title":"Pitfalls in the use of DNA microarray data for diagnostic and prognostic classification","volume":"95","author":"R Simon","year":"2003","journal-title":"J Natl Cancer Inst"},{"issue":"9","key":"pone.0357585.ref036","doi-asserted-by":"crossref","first-page":"1599","DOI":"10.1038\/sj.bjc.6601326","article-title":"Diagnostic and prognostic prediction using gene expression profiles in high-dimensional microarray data","volume":"89","author":"R Simon","year":"2003","journal-title":"Br J Cancer"},{"issue":"9458","key":"pone.0357585.ref037","doi-asserted-by":"crossref","first-page":"488","DOI":"10.1016\/S0140-6736(05)17866-0","article-title":"Prediction of cancer outcome with microarrays: a multiple random validation strategy","volume":"365","author":"S Michiels","year":"2005","journal-title":"Lancet"},{"issue":"750","key":"pone.0357585.ref038","article-title":"High-throughput mass spectrometry maps the sepsis plasma proteome and differences in patient response","volume":"16","author":"Y Mi","year":"2024","journal-title":"Sci Transl Med"},{"issue":"9","key":"pone.0357585.ref039","doi-asserted-by":"crossref","first-page":"732","DOI":"10.1089\/sur.2019.363","article-title":"Staging and Personalized Intervention for Infection and Sepsis","volume":"21","author":"N Beckmann","year":"2020","journal-title":"Surg Infect (Larchmt)"},{"issue":"2","key":"pone.0357585.ref040","doi-asserted-by":"crossref","first-page":"129","DOI":"10.1093\/jpids\/pix021","article-title":"Validation of the Sepsis MetaScore for Diagnosis of Neonatal Sepsis","volume":"7","author":"TE Sweeney","year":"2018","journal-title":"J Pediatric Infect Dis Soc"}],"container-title":["PLOS One"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pone.0357585","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,9,3]],"date-time":"2026-09-03T17:29:22Z","timestamp":1788456562000},"score":1,"resource":{"primary":{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pone.0357585"}},"subtitle":[],"editor":[{"given":"Vijayalakshmi","family":"Kakulapati","sequence":"first","affiliation":[{"name":"Sreenidhi Institute of Science and Technology","place":["INDIA"]}],"role":[{"vocabulary":"crossref","role":"editor"}]}],"short-title":[],"issued":{"date-parts":[[2026,9,3]]},"references-count":40,"journal-issue":{"issue":"9","published-online":{"date-parts":[[2026,9,3]]}},"URL":"https:\/\/doi.org\/10.1371\/journal.pone.0357585","relation":{},"ISSN":["1932-6203"],"issn-type":[{"value":"1932-6203","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,9,3]]}}}