PROJECT SUMMARY Cancer is a group of heterogeneous diseases with diverse clinical characteristics and molecular profiles. Decades of research have discovered many germline genetic risk variants that confer hereditary cancer susceptibilities, and somatic driver mutations that promote malignant transformations of body cells. Because somatic mutations arise in the background of an individual’s germline genome, pre-existing germline variants may influence which synergistic somatic mutations are needed to drive tumorigenesis. Thus, these two groups of variants and their interactions together affect a tumor’s clinical phenotype, such as histological subtype, metastasis sites, and response to treatments. However, as current studies mostly examine germline and somatic variants independently, our knowledge of germline-somatic interplays (GxS) is very limited. To address this knowledge gap and to facilitate precision oncology, we propose a novel computational framework, Variants Interacting in Germline and Soma (VIGAS) that jointly assess the functional impact of these two groups of variants. Using multiple myeloma (MM) as the disease platform, we will develop a suite of bioinformatics methods to discover and characterize GxS via integrative analysis of multi-omics data. We will examine various types of genetic variants in coding and noncoding regions. Unlike existing methods that use simple metrics based on co-occurrences of germline and somatic variants, the VIGAS methods aim to discover GxS with biological significance. We will pursue four specific aims. Aim 1: we will develop a VIGAS-e method to discover GxS based on evolutionary selection, in which germline variants modify selective advantages of somatically mutated genes. Aim 2: we will develop a VIGAS-t method to discover GxS based on transcriptional regulation, in which somatic mutations aggravate pre-existing transcriptional aberrations caused by germline variants. Aim 3: we will combine VIGAS-e and VIGAS-t results to identify GxS that converge on common genes and pathways and are associated with clinicopathological features of MM. We will perform experimental validations of top GxS. Aim 4: we will apply VIGAS to various cancer types. Comparisons across cancer types will reveal associations of GxS with tissue specificity, as well as the risk of developing cancers in different organs. We will implement VIGAS as open-source cross-platform tools, release comprehensive annotations of GxS in an online database, and organize a user community on GitHub. The VIGAS methods and the characterized GxS will greatly improve our understanding of the complex genotype-phenotype relationships in tumorigenesis. Our study will reveal cancer heterogeneities rooted in germline genomes and build the foundation for precision cancer management tailored to individual patients. This will transform cancer management from reactionary approaches toward more proactive approaches.

PROJECT NARRATIVE We will develop novel computational methods to investigate how an individual’s inherited and acquired genetic variants interact to affect the clinical courses of cancers. By applying these methods to various cancer types, we will produce a comprehensive map of synergistic and antagonistic genetic variants that can be leveraged to advance precision cancer management. We will release open-source tools to generate data-driven cancer models for future basic and translational research.