Integrative analysis of genetic association data for individual genome interpretation and translational biological research

Genomics aims to transform the development of new medicines using the power of genomics. Our in-house research engine contains the largest collection of human genotype and phenotype data available, and we are continuously seeking to expand it with the aim to understand the nature and extent of pleiotropy in the human genome and its impact in human health and disease.
One area of focus for Genomics is the generation of polygenic risk scores (PRSs) to inform personalised treatment and screening pathways. It is a widely understood issue that PRS performance in individuals with non-European genetic ancestries is lower than that in individuals with European genetic ancestries, which then impacts the utility of Integrated Risk Tools (IRTs) that combine genetic and non-genetic factors PRSs to generate a measure of absolute disease risk. Our previously published work has already investigated this issue (see Weale et al. 2021, https://doi.org/10.1016/j.amjcard.2021.02.032).
To address these issues further, we intend to use these GWAS data to improve the performance of our PRS and subsequently our IRTs by combining with data that are already available to us.
We also plan to combine these data with other in-house resources, to better detail the mechanistic link between genetic and health outcomes. This research will improve the identification and characterisation of novel drug targets. Bespoke analyses within patient subgroups stratified based on major risk alleles or combinations of different endpoints will enable better definition of the patient populations that would benefit from therapeutic interventions, and more precise evaluation of potential safety concerns.
We will also use this individual level data to explore training and testing IRTs (using independent subsets of the data), which combine PRS with the individual level phenotypes available within ADSP including cognition scores, fluid biomarkers and neuroimaging data.
Finally, for individuals within the randomised control trial cohorts we will explore whether constructed PRS and IRTs are associated with drug response using the available phenotype measures.

Genetic data has the potential to transform health care, by better predicting diseases, understanding and recommending treatment options. This is particularly critical for Alzheimer’s disease, which greatly benefits from early detection.
However, much of the data underpinning this revolution have been generated in individuals of European descent. These data cannot be readily used in non-European populations, which limits broad adoption and widens an already existing gap in access to modern genomic technologies. With this project we are interested in assessing how these genetic predictors can be used in individuals of diverse ancestry. We want to improve the design of these tools so that they become more precise, and usable by a wider range of individuals.