Project 452392

Genetic studies of complex and heritable traits advance our understanding of health and disease. Genome-wide association studies have delivered significant insights into the genetic determinants of complex traits in the past decade. However, due to unresolved analytic challenges, more than two thirds of these earlier studies have excluded the important X chromosome (Xchr; two copies in females like the rest of the genome, but only one copy in males), which accounts for >5% of the genome. This is not optimal as demonstrated by our work on intestinal obstruction in Cystic Fibrosis, where the top disease association signal resides on the Xchr (Sun et al. 2012, Nature Genetics). In response, we will develop new and powerful methods that can analyze the Xchr in different study settings, including understanding how environment modifies genetic effects, and how to predict the risk of a disease. In addition to developing these urgently needed association methods, our research also pays special attention to quality control of the Xchr data, because data-collection tools applied to the Xchr often use those developed for the rest of the genome, which can lead to errors in data. We thus also propose to develop new methods to identify errors in data prior to analyses, and make our association methods robust to data errors. We will test and apply our methods to large-scale genetic datasets, including the 1000 Genomes Project (thousands of individuals from 26 populations) and the UK Biobank (~0.5M individuals and >1000 complex traits and diseases), and improve our understanding of some of the mechanisms that result in sex differences in complex traits and diseases. Finally, as part of the outputs from this research program, we will implement the proposed methods as user-friendly and open resource computational programs so that others can use these state-of-the-art methods to further their studies, develop new screening and treatments, ultimately improving public health.