Background As a common cancer-related death worldwide, pancreatic adenocarcinoma (PAAD) has significantly increased mortality in recent years. In recent years, tumor mutation burden (TMB) has been regarded as the most popular biomarker for PAAD immunotherapy. However, it remains unclear which gene mutations affect TMB and immune response in pancreatic adenocarcinoma. Methods The somatic mutation images of PAAD samples were downloaded from The Cancer Genome Atlas (TCGA) and the International Cancer Genome Consortium (ICGC). Based on the expression data of the TCGA and IGCC cohorts, various bioinformatics algorithms are used for evaluating the prognostic value and functional annotation of some frequently somatically mutated genes. Furthermore, the correlation between gene mutation and tumor immunity was also evaluated. Results The results showed that lysine methyltransferase 2C (KMT2C) and paternally expressed 3 (PEG3) are frequently mutated genes in PAAD. Patients with KMT2C and PEG3 mutations have higher TMB severity and a lousy prognosis. In addition, the mutations of KMT2C and PEG3 genes positively regulate the metabolic and protein-related pathways in PAAD. Meanwhile, significant differences in the composition of the immune cells were observed for KMT2C and PEG3 mutations PAAD patients, for providing additional guidelines for antitumor treatments in various KMT2C and PEG3 mutation states in PAAD. Conclusion This study reveals that KMT2C and PEG3 mutation may serve as biomarkers for predicting prognosis and guiding anti-PAAD immunotherapy for PAAD patients.