Genomic and Transcriptomic Analysis of Breast and Ovarian Cancers

摘要

ABSTRACT Ovarian and breast cancers share common genetic and lifestyle/environmental factors. GWAS have identified more than a hundred genomic regions containing common variants associated with risks of these cancers, several of which confer risks to both cancers. The molecular features of the aggressive subtypes of these cancers--high grade serous ovarian cancer (HGSOC) and estrogen receptor (ER) negative breast cancer—are also remarkably similar, suggesting common genetic and biological mechanisms driving disease development. In the current proposal, we aim to focus on pleiotropic mechanisms underlying susceptibility to ovarian and breast cancer, through tissue-specific transcription-wide analysis of gene expression associated with common variant risk alleles for ovarian and breast cancer identified by GWAS. We will apply a new gene-based association method, PrediXcan, to test the molecular mechanisms through which genetic variation affects ovarian and breast cancer development. The proposed integration of germline genetic data with annotation of whole genome transcription in the relevant tissue types effectively reduces the multiple testing burden faced by GWAS by grouping together multiple risk loci at the gene-level, and further simplifies additional characterization of implicated pathways. We hypothesize that the biological relevance of predictors provided by PrediXcan will allow us to overcome tumor heterogeneity and identify novel genes/pathways for ovarian and breast cancer. We then propose performing functional analyses in experimental models of breast and ovarian cancer to validate the genes and pathways we identify using PrediXcan. This proposal incorporates novel methodology integrating multiple sources of genomic and transcriptomic data to identify the role of genetically regulated gene expression traits in the pathogenesis of ovarian and breast cancers. Mortality is highest for the specific subtypes of these cancers that we will focus on, since the mechanisms underlying the pathogenesis of these cancer subtypes are poorly understood. Genotype and phenotype data from the GAME-ON, OCAC and BCAC consortia, along with the publicly available datasets, TCGA, METABRIC and GTEx, represent a unique opportunity to apply the PrediXcan approach in a large- scale for a well-defined group of cases and controls with high quality epidemiologic data resources .