Machine learning identifies candidate sepsis molecular subtypes and associated markers
Source check: Source checked. Not independent scientific confirmation.
Finding: Transcriptomic clustering and machine-learning feature selection identified three candidate sepsis groups with differing observed 28-day outcomes and a four-gene classifier. A separate dataset and cell experiments provide exploratory checks. The authors caution that these groups are not established clinical phenotypes and need prospective validation.
Technology: Nonnegative-matrix-factorization clustering defined transcriptomic groups; feature-selection methods and an SVM identified and classified subtype-associated markers.
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
- AI contribution
- Nonnegative-matrix-factorization clustering defined transcriptomic groups; feature-selection methods and an SVM identified and classified subtype-associated markers.
- Test and result
- The classifier was assessed in a separate transcriptomic cohort; exploratory cell work examined one marker.
- Limitation / still unknown
- Subtype labels are not established clinical phenotypes; analysis and cell follow-up do not demonstrate clinical utility or treatment efficacy.
Evidence sources and annotation context
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
Evidence sources: https://pmc.ncbi.nlm.nih.gov/articles/PMC13778446/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review date: 22 Nov 2026. Review scope: Primary full-text review as recorded in case evidence; record-level claim and attribution review..
No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.