Identification of common diagnostic biomarkers and immune landscapes in sepsis and acute kidney injury: a transcriptomic study integrating machine learning and single-cell analysis.

Background: Sepsis and acute kidney injury (AKI) are life-threatening conditions often coexisting as sepsis-associated AKI (S-AKI). However, their shared molecular mechanisms and immune heterogeneity remain unclear. This study aims to identify robust diagnostic biomarkers applicable to both conditions and to elucidate their diverse immune microenvironments using integrated transcriptomic approaches. Methods: Transcriptomic datasets for sepsis and AKI were analyzed, with multiple cohorts used for training and external validation. Differential gene expression and WGCNA identified key modules, while LASSO and Random Forest algorithms screened shared hub genes. A diagnostic nomogram was constructed and evaluated using ROC and decision curve analyses. Single-cell RNA sequencing data were further analyzed to determine cellular localization, functional pathways, and intercellular communication. Results: Four hub genes ( Conclusion: We identified a reliable four-gene diagnostic signature shared by sepsis and AKI and characterized their shared immune heterogeneity and inferred intercellular communication networks. These findings provide potential targets for early diagnosis and therapeutic intervention in sepsis-associated renal injury.