Two Gene Set Variation Index as Biomarker of Bacterial and Fungal Sepsis

Abstract

Background. The incidence of sepsis has been increasing in recent years. The molecular mechanism of different pathogenic sepsis remains elusive, and biomarkers of sepsis against different pathogens are still lacking. Methods. The microarray data of bacterial sepsis, fungal sepsis, and mock-treated samples were applied to perform differentially expressed gene (DEG) analysis to identify a bacterial sepsis-specific gene set and a fungal sepsis-specific gene set. Functional enrichment analysis was used to explore the body’s response to bacterial sepsis and fungal sepsis. Gene set variation analysis (GSVA) was used to score individual samples against the two pathogen-specific gene sets, and each sample gets a GSVA index. Receiver operating characteristic (ROC) curve analysis was performed to evaluate the diagnostic value of sepsis. An independent data set was used to validate the bacterial sepsis-specific GSVA index. Results. The genes differentially expressed only in bacterial sepsis and the genes differentially expressed only in fungal sepsis were significantly involved in different biological processes (BPs) and pathways. This indicated that the body’s responses to fungal sepsis and bacterial sepsis are varied. Twenty-two genes were identified as bacterial sepsis-specific genes and upregulated in bacterial sepsis, and 23 genes were identified as fungal sepsis-specific genes and upregulated in fungal sepsis. ROC curve analysis showed that both of the two pathogen sepsis-specific GSVA indexes may be a reliable biomarker for corresponding pathogen-induced sepsis (AUC=1.000), while the mRNA of CALCA (also known as PCT) have a poor diagnostic value with AUC=0.512 in bacterial sepsis and AUC=0.705 in fungi sepsis. In addition, the AUC of the bacterial sepsis-specific GSVA index in the independent data set was 0.762. Conclusion. We proposed a bacterial sepsis-specific gene set and a fungal sepsis-specific gene set; the bacterial sepsis GSVA index may be a reliable biomarker for bacterial sepsis.