Genome-wide association studies (GWAS) do not pinpoint the disease-causal variant, and genomics experiments do not pinpoint which molecular events influence disease risk. What is needed are methods that combine GWAS and functional genomics to infer disease risk-altering causal variants. We aim to identify the variants that have a causal effect on trait via their effect on the binding of the Vitamin D Receptor (VDR). Two-sample Mendelian randomization is limited by unobserved pleiotropic effects of candidate variants which may explain trait effects independently of VDR binding. This project addresses this problem by adding the cis-acting DNA variants that explain Vitamin D (25OHD) Serum level in specific tissues to the model. This allows us to infer variants that are causal for physiological/disease trait variation via their effect on VDR binding.