Quantitative Methods for Genetic Epidemiology

Project Summary The advancements of genomic technologies and assemblies of large disparate sets of biological and health data have outpaced the ability to integrate these different sources of information. Powerful statistical methods and software are needed to fill this gap in order to provide novel understandings of biological processes, as well as provide better predictions of human diseases to achieve the vision of personalized medicine. The broad goals of this project are to advance genetic epidemiology studies of human traits and diseases by expanding our development of statistical analytic methods and software encompassing four main areas: 1) multivariate methods to decipher genetic contributions; 2) statistical fine-mapping of genetic variants; 3) causal mediation methods; 4) polygenic risk scores (PRS) for predicting disease. Although these areas might appear broad and disparate, there is pressing need to build more integrative methods across these domains. For example, because molecular pleiotropy is pervasive, multivariate analysis is essential to identify shared genetic factors acting through common biological mechanisms of multiple traits, and when using PRS to predict disease, complex traits are often better predicted when multivariate correlated traits are used. And, the methods used for statistical fine-mapping, including use of annotation, are relevant for creating PRS to predict disease. Our team, involving statistical geneticists, computational biologists, genetic epidemiologist and clinical investigators, has decades of experience and will capitalize on the extensive resources and collaborations we have developed. Our novel methods will be applied to a broad range of diseases, with ultimate aims to better understand disease etiology and improved disease prediction across different ethnic groups to reduce health disparities. User-friendly software will be distributed with open access to the scientific community. We will take advantage of rapidly evolving technologies, biologic and computational insights from multiple fields, and evolving public health and clinical unmet needs to inform our science.