Sepsis is a systemic inflammatory response syndrome, which is mostly induced by infection in the lungs, the abdomen and the urinary tract. The present study is aimed to investigate the mechanisms of sepsis. Expression profile of E‑MTAB‑4421 (including leukocytes isolated from 207 survived and 58 non‑survived patients with sepsis) and E‑MTAB‑4451 (including leukocytes isolated from 56 survived and 50 non‑survived patients with sepsis) were downloaded from the European Bioinformatics Institute database. Based on the E‑MTAB‑4421 expression profile, several differentially expressed genes (DEGs) were identified and performed with hierarchical clustering analysis by the limma and pheatmap packages in R. Using the BioGRID database and Cytoscape software, a protein‑protein interaction (PPI) network was constructed for the DEGs. Furthermore, module division and module annotation separately were conducted by the Mcode and BiNGO plugins in Cytoscape software. Additionally, the support vector machine (SVM) classifier was constructed by the SVM function of e1071 package in R, and then verified using the dataset of E‑MTAB‑4451. A total of 384 DEGs were screened in the survival group. The PPI network was divided into 4 modules (modules A, B, C and D) involving 11 DEGs including microtubule‑associated protein 1 light chain 3 alpha (MAP1LC3A), protein kinase C‑alpha (PRKCA), metastasis associated 1 family member 3 (MTA3), and scribbled planar cell polarity protein (SCRIB). SCRIB and PRKCA in module B, as well as MAP1LC3A and MTA3 in module D, might function in sepsis through PPIs. Functional enrichment demonstrated that MAP1LC3A in module D was enriched in autophagy vacuole assembly. Finally, the SVM classifier could correctly and effectively identify the samples in E‑MTAB‑4451. In conclusion, DEGs such as MAP1LC3A, PRKCA, MTA3 and SCRIB may be implicated in the progression of sepsis, and need further and more thorough confirmation.