Genetic epidemiology of endometrial cancer

Suitable for PhD students only.

BACKGROUND

Endometrial cancer is the most commonly diagnosed invasive gynaecological cancer in developed countries. In contrast with many cancers, the incidence and mortality of endometrial cancer is steadily increasing, largely due to increasing rates of obesity, the strongest risk factor for this disease. Through leadership of the Endometrial Cancer Association Consortium (ECAC), our lab runs the largest genetic study of endometrial cancer. To date, we have identified 16 genetic regions associated with endometrial cancer predisposition by genome-wide association study (GWAS), which account for ~25% of the genetic heritability attributable to common genetic variants (O’Mara et al, Nat Commun 2018). Incorporation of existing GWAS data with newly acquired GWAS datasets from international collaborators will identify further genetic regions associated with endometrial cancer risk. Additionally, we have approved access to large, well-phenotyped international datasets (e.g., UK Biobank, N = 500,000). This allows us unparalleled ability to examine the genetics of endometrial cancer, as well as explore its relationship with risk factors, such as obesity.

AIMS

To identify new genetic risk regions for endometrial cancer, by performing the largest GWAS meta-analysis for this disease. To use computational approaches to identify and explore risk factors of endometrial cancer. To use genetic data to construct and test risk prediction models for endometrial cancer.

APPROACHES

This project will use standard GWAS pipelines to identify genetic variants associated with endometrial cancer risk, including imputation, QC and association testing. Post-GWAS analyses to explore novel regions could also be performed (e.g., eQTL analyses, integration with functional genomic datasets). The relationship between endometrial cancer and potential/known risk factors will be performed using approaches such as genetic correlation (LD Score Regression) and Mendelian randomization. Endometrial cancer risk prediction models will be constructed using polygenic risk scores in combination with endometrial cancer environmental risk factors and tested for efficacy in independent datasets.