Incorporating WGCNA and Machine Learning to Identify ADAP2 as a Critical Efferocytosis-Related Gene in Sepsis.

backgroundSepsis, a life-threatening organ dysfunction caused by dysregulated host responses to infection, frequently involves impaired macrophage efferocytosis that leads to apoptotic cell accumulation, secondary necrosis, and persistent inflammation. Early prognostic stratification remains challenging, as current biomarkers lack sufficient specificity and sensitivity, underscoring the urgent need for novel prognosis-related indicators.

methodsWe integrated bulk transcriptomic data from a discovery cohort (GSE205672) and an independent validation cohort (GSE133822) with single-cell RNA-seq profiles of early- and late-stage sepsis (GSE167363, GSE175453). WGCNA and five consensus machine-learning algorithms were combined to screen core efferocytosis-associated genes, and expression was validated via qPCR in PBMCs from sepsis patients and CLP-induced septic mice.

resultsADAP2 was identified as the core gene achieving strict consensus across all five algorithms, with early upregulation and late depletion in sepsis, predominant expression in monocytes/macrophages-particularly M1-like and IFN-responsive subsets-and a significant correlation with efferocytosis scores and immune cell infiltration. Its expression was negatively correlated with sepsis severity (SOFA score) and showed a trend toward worse survival in patients with low ADAP2 levels.

conclusionsThis multi-dimensional transcriptomic study establishes ADAP2 as a candidate biomarker with potential prognostic value in sepsis, closely linked to macrophage efferocytosis. These findings may aid early risk stratification and inform macrophage-directed immunotherapies, although prospective validation and functional studies are required.