Skip to navigation Skip to main content Skip to footer

Approved research

Development of cross-phenotype meta-analysis methods that can account for effect size heterogeneity between diseases and sample overlap

Principal Investigator: Professor Buhm Han
Approved Research ID: 46263
Approval date: April 23rd 2019

Lay summary

Genome-wide association studies (GWAS) are a method to find genes that increase susceptibility to diseases. To date, many studies used GWAS to find which genes are causing which diseases, which have tremendously helped our understanding of diseases. Surprisingly, recently, many genes turned out to have effects on multiple different diseases. This phenomenon that the same gene (or variant) has effects on multiple phenotypes is called 'pleiotropy'. Pleiotropy is prevalent among diseases in the same category, such as among autoimmune diseases, or among psychiatric disorders. If pleiotropic loci are found, they can suggest that the diseases are genetically similar. Pleiotropic loci can also suggest that a drug can be effective on multiple diseases. Thus, finding pleiotropic loci will be beneficial to our understanding of diseases. However, how to effectively find pleiotropic loci is unclear. Sometimes, the same gene can have differing effects on two diseases - it may have small effect on disease A, and large effect on disease B. This phenomenon is called 'effect size heterogeneity'. Moreover, the same healthy individuals may have been used for analyses of multiple diseases, which is a situation called 'sample overlap'. Both heterogeneity and sample overlap can be a challenge when we try to analyze multiple diseases together to find pleiotropic loci. In this project, we will develop a method that can address these challenges and can find pleiotropic loci very well. We will use the UK Biobank dataset as a test dataset for this method.