Integration of polygenic risk scores and clinical data into an open platform that enables pragmatic trials

Research into the genetic basis of human diseases has led to powerful methods for polygenic risk scores (PRS) and the key challenge now shifts to application of these knowledge to improve patient care. We propose to develop methods and platforms to evaluate the clinical benefits of PRS. First, we will evaluate the clinical benefit of PRS by: (i) Evaluating early intervention benefits at population level.

We plan to evaluate the benefits of PRS-guided intervention for common diseases using population-level EHR data (e.g. CVD- COVID-UK). (ii) Constructing polygenic scores for treatment responses.

We propose a novel imputation method to construct more powerful pharmacogenomic PRS. (iii) Identifying disease subtypes with varying treatment responses. Second, we will integrate PRS with clinical data by: (i) Incorporating EHR data to improve prediction accuracy.

We aim to construct better risk models through integrating PRS and diagnostic/prescription data. (ii) Leveraging PRS to construct metabolite/proteomics-based risk scores for complex diseases.

Third, we will implement PRS computation infrastructure in NHS systems by: (i) Calibrating PRS to the NHS covered populations.

We will develop PRS calibration and test them on routinely collected genotype data through NHS GLH. (ii) Implementing transferable PRS tools that are compatible with medical management systems.