Establishment of a 10-gene prognostic model for gastric cancer based on the tumor immune microenvironment

Abstract

Gastric cancer seriously affects the health of modern people. The immune microenvironment of gastric cancer tissue is key to gastric cancer progression. We downloaded training and validation sets data from The Cancer Genome Atlas and Gene Expression Omnibus. Single-sample gene set enrichment analysis was used to sort patients into high, middle, and low immunity groups, of which immune infiltration in the high immunity group was substantially higher than of other two groups. Genes in high and low immunity groups expressed prominent differences. Further, the enrichment of differentially expressed genes was found mainly in immune-related pathways. Subsequently, an immune-related prognostic model was established, composed of ten prognosis-related genes identified by univariate risk regression, least absolute shrinkage and selection operator Cox, and multivariate risk regression. Survival analysis and receiver operating characteristic curves suggested good diagnostic efficacy of this model, and feature genes were linked to the degree of immune infiltration. An independent test suggested that the risk score could independently determine patient outcomes. We combined all clinical information and risk scores to establish a nomogram that could predict patient's prognosis. A prognostic model composed of 10 prognosis-related genes was generated with good diagnostic efficacy in predicting prognoses of gastric cancer patients.