Polygenic Contributions To Genetic Susceptibility: Improving Infrastructure For Understanding Liability To Disease

Abstract

to infrastructure that can be used to study liability to disease. Two main aims are pursued: enhancing infrastructure, and improving understanding of complex disease under the lens of shared polygenic contributions in commonly comorbid conditions involving pain. The first aim focuses on enhancing the approach to calculating polygenic contributions to genetic susceptibility by improving the Polygenic Risk Score Continuous Shrinkage (PRS-CS) method. Using benchmarking analyses, the proposed improvements are validated with the latest Genome-Wide Association Study (GWAS) summary statistics from 8 traits that vary by GWAS sample size and expected polygenicity. Performance measures, including time, accuracy, computational efficiency/accessibility, and correlation with traits in the BioVU electronic health records, are compared against the PRS-CS python package. The second aim focuses on improving genetic susceptibility to complex traits related to pain from three angles. In the first of these, the polygenic risk of pain in 8 parts of the body are used to derive a shared component and used to understand how the strength of this component interacts with demographic features such as age, sex, and BMI in the prediction of chronic overlapping pain conditions. In the second part of this aim, genetic susceptibility to common comorbidities of Fibromyalgia across different disease categories, like auto-immune and psychiatric conditions are used to predict Fibromyalgia, and the largest Fibromyalgia GWAS of its time was performed and used to identify a novel genome-wide significant variant. In the third part of this aim, the polygenic contributions and comorbidities of opioid use disorder are explored. In conclusion, this thesis presents advancements in understanding polygenic genetic susceptibility to pain-related phenotypes, as well as infrastructure for calculating and applying polygenic risk scores.