Construction of a novel immune-related prognostic-predicting model of gastric cancer

Gastric cancer (GC) is a common primary stomach tumor of the central nervous system with a poor prognosis. In this study, 274 differentially expressed immune-related genes (DEIRGs) were identified among six cell sub -populations in GSE112302 single-cell RNA sequencing (scRNA-seq) data of GC. Those DEIRGs were able to divide GC patients into three distinct subtypes with different overall survivals and tumor microenvironment. By uni-variate Cox and LASSO regression analyses, eight immune-related genes, including CTGF, CXCL3, CXCR4, NRP1, OAS1, SP1, STC1 and TAP1, were identified as GC prognostic signatures. Accordingly, a risk score model for predicting GC prognosis was constructed in TCGA-GC training cohort and validated in the external GSE66229 dataset. Moreover, a nomogram for predicting the survival of GC patients was also established based on inde-pendent prognostic factors (age, grade, cancer status and risk score) identified by univariate and multivariate Cox regression analyses. In addition, Gene Set Variation Analysis (GSVA) analysis indicated that the prognostic immune signatures may regulate GC via inflammation and cell proliferation related pathways, such as DNA replication, complement and coagulation cascades, focal adhesion and TGF-beta signaling pathway.

Automated summary

274 differentially expressed immune-related genes were identified in gastric cancer, and a risk score model and nomogram for predicting prognosis were established. The immune signatures may regulate gastric cancer through inflammation and cell proliferation related pathways.