In conclusion, we presented a novel gene-based test, PrediXcan that incorporates functional information with regard to gene regulation to identify genes associated with disease traits in large GWAS or whole genome sequence datasets. Our method has the advantage of providing biological insights into the mechanism, namely regulation of gene expression, and direction of effect. This approach can be readily applied to existing GWAS datasets through the use of our publically available PredictDB resource. We further show the utility of our approach by identifying and replicating a number of novel candidate associations within the previously analyzed WTCCC dataset.
