IDENTIFICATION OF A NOVEL SEPSIS PROGNOSIS MODEL: BASED ON TRANSCRIPTOME AND PROTEOME ANALYSIS

Abstract

Sepsis is a highly prevalent and deadly disease. Currently, there is a lack of ideal biomarker prognostis models for sepsis. We attempt to construct a model capable of predicting the prognosis of sepsis patients by integrating transcriptomic and proteomic data. Through analysis of proteomic and transcriptomic data, we identified 25 differentially expressed genes (DEGs). Single-factor Cox-Lasso regression analysis identified 16 DEGs (overall survival-DEGs) associated with patient prognosis. Through multifactor Cox-Lasso regression analysis, a prognostic model based on these 16 genes was constructed. Kaplan-Meier survival analysis and receiver operating characteristic curve analysis were used to further validate the high stability and good predictive ability of this prognostic model with internal and external data. Gene Ontology and Kyoto Encyclopedia of Genes and Genomes pathway enrichment analysis of overall survival-DEGs and differentially expressed genes between high and low-risk groups based on the prognostic model revealed significant enrichment in immune-related pathways, particularly those associated with viral regulation.