Our underlying model is closely related to the approach developed by Flutre et al. [10], which considers the different but related problem of maximising the power to discover eQTLs in expression datasets of multiple tissues. A key feature of our approach is that it only requires single SNP p-values and their minor allele frequencies (MAFs), or estimated allelic effect and standard error, combined with closed form analytical results that enable quick comparisons, even at the genome-wide scale. Our Bayesian procedure provides intuitive posterior probabilities that can be easily interpreted. A main application of our method is the systematic comparison between a new GWAS dataset and a large catalogue of association studies in order to identify novel shared mechanisms. We demonstrate the value of the method by re-analysing a large scale meta-analysis of blood lipids [15] in combination with a gene expression study in 966 liver samples [16].
