Comprehensive identification of germline-somatic interactions

Cancer remains a significant source of morbidity and mortality worldwide. Advances in next generation
sequencing technologies have allowed extensive profiling of the genetic variants present in tumors and revealed
daunting levels of inter-tumoral heterogeneity. Tumor genomic datasets have been extensively mined to reveal
germline risk variants and to characterize heterogeneous patterns of somatic alteration that drive tumor
progression. However, little attention has been given to interactions between genetic background and somatic
changes, which could represent a major driver of heterogeneity. New evidence suggests that germline-somatic
interactions are prevalent and our preliminary data support that some such interactions directly influence
individual disease risk and potential to respond to therapies. This proposal will develop computational strategies
to identify germline-somatic interactions and to characterize them in the context of molecular and clinical
phenotypes, enabling new understanding of their role in inter-tumoral heterogeneity.
Germline-somatic interactions have been challenging to study due to the limited amount of available data. To
address this challenge, we have compiled tumor genomic data from public sources to boost our sample size to
almost 45,000 tumor whole-exome and whole-genome sequences. Our analysis will focus on three major forces
that shape the tumor genome: (i) the mutational processes that generate somatic mutations, (ii) the molecular
organization of oncogenic pathways which determines the genes that can effectively drive cancer, and (iii) the
immune system which acts as a selective force throughout tumor development. We will focus hypothesis testing
with strict criteria for selecting germline variants and somatically altered genes likely to interact based on
established tumor biology. To identify and characterize germline-somatic interactions we will:
1) Elucidate germline variants affecting the somatic mutational landscapes of human cancers
2) Reveal germline variants that modify somatic activation of hallmark oncogenic pathways
3) Establish the role of pathway-specific variant burden in cancer predisposition, overall survival, and response
to immunotherapy
Our team of co-investigators includes strong complementary expertise in analysis of mutational processes,
genetic variation effects on molecular pathways and immunity, cancer biology, statistical methods and
bioinformatic software dissemination. Careful attention will be given to statistical considerations including power,
controlling false discovery rates, and validation in independent datasets. This proposal will produce A) novel
bioinformatics tools designed specifically to detect and annotate germline-somatic interactions, B) new
understanding of the contribution of germline variation to tumor progression, and C) a set of validated germline-
somatic interactions affecting cancer risk, tumor evolution and immunotherapy response.

Grant Number: 5R01CA269919-04
NIH Institute/Center: NIH
Principal Investigator: Hannah Carter