Decoding Germline Genetic Influence on Cancer Somatic Mutation Acquisition

Abstract

In this paper we investigate the impact of germline variants on the somatic mutation profiles of key cancer drivers, focusing on TP53 and KRAS, in a cohort of 35,914 patients (16,424 TP53-mutated and 6,297 KRAS-mutated). Candidate genes were selected based on protein–protein interaction data from STRING and BioGrid, yielding tailored panels for TP53 (including BRCA1, MDM2, ATM, among others) and KRAS (including RAF1, BRAF, EGFR, etc.). Using a multi-model machine learning framework—incorporating logistic regression, random forests, and feed-forward neural networks—we captured both linear and non-linear associations in high-dimensional SNP data. Our analysis revealed a significant association between BRCA1 variants and TP53 mutation status in breast carcinoma, supporting the role of inherited DNA repair deficiencies in promoting somatic TP53 alterations. In contrast, no single germline driver was identified for KRAS mutations, suggesting a more complex interplay of factors in KRAS-driven oncogenesis. These findings underscore the value of integrating biologically informed candidate gene selection with advanced computational models to enhance our understanding of cancer genomics and inform future risk assessment and therapeutic strategies.