Bioinformatic Analysis of Prognostic and Immune-Related Genes in Pancreatic Cancer

Abstract

Pancreatic cancer (PC) is a malignant tumor with poor prognosis. The poor effect of surgery and chemotherapy makes the research of immunotherapy target molecules significant. Therefore, identifying the new molecular targets of PC is important for patients. In our study, we systematically analyzed molecular correlates of pancreatic cancer by bioinformatic analysis. We characterized differentially expressed analysis based on the TCGA pancreatic cancer dataset. Then, univariate Cox regression was employed to screen out overall survival-(OS-) related DEGs. Based on these genes, we established a risk signature by the multivariate Cox regression model. The ICGC cohort and GSE62452 cohort were used to validate the reliability of the risk signature. The impact of T lymphocyte-related genes from risk signature was confirmed in PC. Here, we observed the correlation between the T lymphocyte-related genes and the expression level of targeted therapy. We established a five-mRNA (LY6D, ANLN, ZNF488, MYEOV, and SCN11A) prognostic risk signature. Next, we identified ANLN and MYEOV that were associated with T lymphocyte infiltrations (P<0.05). High ANLN and MYEOV expression levels had a poorer prognosis in decreased T lymphocyte subgroup in PC. Correlation analysis between ANLN and MYEOV and immunomodulators showed that ANLN and MYEOV may have potential value in pancreatic cancer immunotherapy.