NOVEL STATISTICAL METHODS FOR POLYGENIC RISK SCORE GENERATION IN CARDIOVASCULAR DISEASES

Abstract

Polygenic risk scores (PRS) are relatively novel tools for risk prediction, serving as a quantitative singular value which depicts a patient’s genetic disposition for a certain disease. Given that many current clinical risk predictors do not address heritability within their calculations, PRS are likely to improve prediction, especially in the case of complex diseases which are influenced by a combination of genetic, environmental and lifestyle factors. Altogether, PRS studies have been pursued for their abilities in trait detection, therapeutic intervention and disease protection, with much potential in personalized/precision medicine where each interpretation is unique and based on a patient’s genotype. However, despite the numerous advances over years, PRS have yet to reach the level where they can be implemented into standard clinical practices as originally intended. The goal is to develop PRS which are applicable to global populations, which has yet to be achieved due to the inconsistency and general skepticism regarding the method. Furthermore, PRS have yet to reach the upper threshold for risk prediction, as indicated by the heritability that remains unaccounted for with PRS calculations. Thus, this thesis addresses how PRS can inform and guide clinical decision-making for complex decisions with strong, genetic dispositions. It also presents novel approach to PRS aimed at mitigating some of its current limitations.