Abstract
Sepsis is a disease with a high morbidity and mortality rate. At present, there is a lack of ideal biomarker prognostic models for sepsis and promising studies using prognostic models to predict and guide the clinical use of medications. In this study, 71 differentially expressed genes (DEGs) were obtained by analyzing single-cell RNA sequencing (scRNA-seq) and transcriptome RNA-seq data, and Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment pathway analyses were performed on these genes. Then, a prognosis model with CCL5, HBD, IFR2BP2, LTB, and WFDC1 as prognostic signatures was successfully constructed after univariate LASSO regression analysis and multivariate Cox regression analysis. Kaplan-Meier (K-M) survival analysis, receiver operating characteristic (ROC) time curve analysis, internal validation, and principal component analysis (PCA) further validated the model for its high stability and predictive power. Furthermore, based on a risk prediction model, gene set enrichment analysis (GSEA) showed that multiple cellular functions and immune function signaling pathways were significantly different between the high- and low-risk groups. In-depth analysis of the distribution of immune cells in healthy individuals and sepsis patients using scRNA-seq data revealed immunosuppression in sepsis patients and differences in the abundance of immune cells between the high- and low-risk groups. Finally, the genetic targets of immunosuppression-related drugs were used to accurately predict the potential use of clinical agents in high-risk patients with sepsis.