Comprehensive Analysis of Molecular Subtypes and Hub Genes of Sepsis by Gene Expression Profiles.

Sepsis is a systemic inflammatory response syndrome (SIRS) with heterogeneity of clinical symptoms. Studies further exploring the molecular subtypes of sepsis and elucidating its probable mechanisms are urgently needed. Microarray datasets of peripheral blood in sepsis were downloaded from the Gene Expression Omnibus (GEO) database, and differentially expressed genes (DEGs) were identified. Weighted gene co-expression network analysis (WGCNA) analysis was conducted to screen key module genes. Consensus clustering analysis was carried out to identify distinct sepsis molecular subtypes. Subtype-specific pathways were explored using gene set variation analysis (GSVA). Afterward, we intersected subtype-related, dramatically expressed and module-specific genes to screen consensus DEGs (co-DEGs). Enrichment analysis was carried out to identify key pathways. The least absolute shrinkage and selection operator (LASSO) regression analysis was used for screen potential diagnostic biomarkers. Patients with sepsis were classified into three clusters. GSVA showed these DEGs among different clusters in sepsis were assigned to metabolism, oxidative phosphorylation, autophagy regulation, and VEGF pathways, etc. In addition, we identified 40 co-DEGs and several dysregulated pathways. A diagnostic model with 25-gene signature was proven to be of high value for the diagnosis of sepsis. Genes in the diagnostic model with AUC values more than 0.95 in external datasets were screened as key genes for the diagnosis of sepsis. Finally, ANKRD22, GPR84, GYG1, BLOC1S1, CARD11, NOG, and LRG1 were recognized as critical genes associated with sepsis molecular subtypes. There are remarkable differences in and enriched pathways among different molecular subgroups of sepsis, which may be the key factors leading to heterogeneity of clinical symptoms and prognosis in patients with sepsis. Our current study provides novel diagnostic and therapeutic biomarkers for sepsis molecular subtypes.