Breast cancer is a complex disease with both genetic and non-genetic factors contributing to risk. In addition, many established risk factors for breast cancer including mammographic density, age at menarche, age at natural menopause, height, and body mass index (BMI) are under strong genetic control. We recently found indications that mammographic density and breast cancer are under shared genetic control, highlighting specific causal pathways. Further characterization of such shared genetic origin would provide invaluable insights in breast cancer etiology and biology, but lack of adequate statistical methods and large-scale empirical datasets has precluded such analysis. We here propose to develop new statistical methodology in order to quantify and characterize the overall shared genetic origin between a given breast cancer risk factor and breast cancer. We will then quantify the proportion of the observed shared genetic origin that can be explained by already identified loci. Such analysis will inform about possible disease pathways. We will develop novel statistical methodology based on variance component theory to robustly quantify the genome-wide shared genetic origin (so called cross-trait heritability) between a given breast cancer risk factor phenotype and breast cancer. We will then study what proportion of the overall shared genetic basis can be explained by already known loci. Our method requires only GWAS summary statistics as input, enhancing its applicability to large-scale consortia-based datasets. We will use our method to estimate the shared genetic origin of breast cancer and each of five breast cancer risk factors: mammographic density, age at menarche, age at natural menopause, height, and BMI. We have acquired datasets of European and African American ancestry that will allow us to estimate and compare cross-trait heritability between ethnicities. For women of European ancestry, we have access to GWAS summary statistics based on 15,000 breast cancer cases and 20,000 controls, 5,000 women with mammographic density measurements and 7,000 women for age at menarche, age at natural menopause, height and BMI. For women of African-American Ancestry, we have access to GWAS summary statistics based on 3,000 breast cancer cases and 2,800 controls as well as 2,400 women for BMI and height. This research application describes an innovative and cost-efficient approach to study the causal mechanisms underlying the associations between breast cancer risk factors and breast cancer. We will develop new methodology to quantify the shared genetic origin between two correlated traits. Our method requires only GWAS summary statistics as input, enhancing its applicability to large-scale consortia-based datasets. We will make our methodology publicly available by releasing user-friendly software. Ultimately, characterization of the genetics underlying breast cancer will lead to novel and important insights into breast cancer biology and etiology.
We propose to develop new methodology to estimate the shared genetic origin between two correlated traits. We will then use our method to quantify the shared genetic origin between a given breast cancer risk factor and breast cancer in populations of European and African-American ancestry. Such research will increase our understanding about the biological mechanisms underlying breast cancer and point towards specific causal pathways involved in breast cancer etiology.
| Lindström, Sara; Finucane, Hilary; Bulik-Sullivan, Brendan et al. (2017) Quantifying the Genetic Correlation between Multiple Cancer Types. Cancer Epidemiol Biomarkers Prev 26:1427-1435 |
| Vilhjálmsson, Bjarni J; Yang, Jian; Finucane, Hilary K et al. (2015) Modeling Linkage Disequilibrium Increases Accuracy of Polygenic Risk Scores. Am J Hum Genet 97:576-92 |
| Bulik-Sullivan, Brendan; Finucane, Hilary K; Anttila, Verneri et al. (2015) An atlas of genetic correlations across human diseases and traits. Nat Genet 47:1236-41 |
| Finucane, Hilary K; Bulik-Sullivan, Brendan; Gusev, Alexander et al. (2015) Partitioning heritability by functional annotation using genome-wide association summary statistics. Nat Genet 47:1228-35 |
| Bulik-Sullivan, Brendan K; Loh, Po-Ru; Finucane, Hilary K et al. (2015) LD Score regression distinguishes confounding from polygenicity in genome-wide association studies. Nat Genet 47:291-5 |
| Lindström, Sara; Thompson, Deborah J; Paterson, Andrew D et al. (2014) Genome-wide association study identifies multiple loci associated with both mammographic density and breast cancer risk. Nat Commun 5:5303 |
