Developing a Bayesian methodology for identifying disease-associated dysregulation of genes and gene subnetworks by integration of GWAS and eQTL data, with application to cardiovascular diseases.

We aim to develop a Bayesian methodology for identifying disease-associated dysregulation of genes and gene subnetworks by integrating GWAS and eQTL data. This will improve our understanding of the underlying mechanism of the common, non-infectious diseases. The genes and gene modules will represent prime targets for pharmacological intervention. We plan to use phenotypes related to cardiovascular diseases (CVD) to validate our method. Our software will produce ranked list of genes and gene modules which are statistically associated to diseases. These will have clinical utility for prognosis or treatment. The proposed method will also improve the prediction of disease propensity of any individual given his genotype. Hence it can be used for improved diagnosis and prevention of common, non-infectious diseases. We have formulated the theory for the proposed method, and we are in the process of developing a software tool. We will find CVD cases and matched controls from the UK Biobank data, and apply our method to discover genes and gene subnetworks whose dysregulation is statistically coupled to increased risk for CVD. We would like to use the largest sample size available to maximize the power of detecting statistical associations.