About
The unique genetic profile of an individual can be used together with the environmental factors such as life-style to predict a particular disease risk. There are lots of literatures involving building up statistical models to extract useful information for disease risk prediction from genetic profile. However, most are only dealing with European population. Compared with European genetic studies, non-European studies usually have smaller sample size. What's more, the genetic profiles are different for a large proportion for Europeans and non-Europeans. So the existing disease risk prediction methods using European genetic information do not perform well when predicting disease risks for non-Europeans.
In this project, we aim to (1) propose new statistical methods to do disease risk prediction through extracting useful genetic information from non-European populations; (2) For a list of diseases (such as Type 2 diabetes, cardiovascular, lupus and inflammatory diseases) which are observed in UK Biobank cohort, we would calculate the disease risk using genetic profile in UK Biobank through applying currently existing methods, then calculate the disease risk using genetic profile from a non-European biobank with applying our proposed method, and finally compare the risk prediction performance in the two populations. Our goal is to reduce the gap in the accuracy of disease risk prediction using genetic profile between European and non-European populations. This project will pave the way for disease diagnostics for non-Europeans and further contribute to precision medicine. This project is intended to last for 3 years.