Sepsis is a fatal whole-body inflammatory response that complicates a serious infection. To elucidate the molecular mechanism of sepsis, transcription profile data of GSE12624 which included a total of 70 samples (34 sepsis samples and 36 non-sepsis samples) was downloaded. The t test based on Bayes method in limma package was used to identify differentially expressed genes (DEGs) between sepsis and non-sepsis samples (criterion: P value <0.05). Gene Ontology (GO) enrichment analysis was conducted to investigate the biological processes involved DEGs. Protein-protein interaction (PPI) network and sub-network analysis were conducted to investigate the interactions between DEGs. A total of 894 DEGs, including 479 downregulated DEGs and 415 upregulated DEGs, were identified in sepsis samples comparing with non-sepsis samples. GO enrichment analysis showed that DEGs mainly involved in cellular metabolic process, primary metabolic process, and response to organic cyclic compound. In the PPI network, four genes of CDC2, GTF2F2, PCNA, and SMAD4 with degrees more than 10 were identified. Subsequently, four sub-networks, in which genes of PTBP1, PSMA3, PSMA6, PSMB9, PSMB10, and GADD45 had relative high degrees were identified from the PPI network. After the discussion referring to previous studies, we suggested that CDC2, GTF2F2, PCNA, SMAD4 PSMA3, PTBP1, and GADD45 might be used as new therapeutic targets for sepsis. However, experiments should be further performed to prove the practical utility of these candidates.