Approved Research

Portability of Genetic Risk Scores to Underrepresented Populations

Lay summary

Throughout modern history, genetic study designs have been heavily influenced and shaped by individuals of European ancestry. This has, in particular, been exacerbated over the last century with advances in genomics: while this group represents only 16% of the global population, a 2019 study showed that around 78% of data used in genome-wide association studies (GWAS) originated from people of European descent. As a result, genomics and medicine have been biased by the default test subject - males of Northern European origin. Even though models for prediction of genetic risk may be present for underrepresented populations, they face risk of being disregarded by clinicians due to their lower accuracy.

The goal of this research is to port genetic risk prediction models for some of the most common diseases, trained in European populations, to Africans. We build on our track record of large-scale implementation of polygenic risk scores, collaborations with the Breast Cancer Association Consortium (BCAC) and our involvement in developing polygenic risk scores for lipid traits in African populations. Thus, we propose the following objectives:

1. a) To gather a training dataset of African populations from public resources (1000 Genomes Project, UK Biobank) as well as existing collaborations (H3Africa, Million Veteran Program).
2. b) To pilot on portability to Africans of existing breast cancer polygenic risk scores.
3. c) To extend our piloted model for portability to diabetes type 2, coronary artery disease, atrial fibrillation and inflammatory bowel disease. Together with breast cancer, these five traits affect ~20% of mortality cases in industrialised nations.

According to the Global Alliance for Genomics and Health, access to benefits of genome research for all peoples is not only a moral imperative but also a question of justice. Research conducted for this work will have an impact in our ability to translate European-trained models for prediction of genetic risk not only for African but for other underrepresented populations. Our research outputs will make the fruits of human genome research more equitable, diverse and inclusive. The project will last for 36 months and a PhD student will be working full time on this.