Abstract

Diseases often co-occur in individuals more often than expected by chance, and this may be explained by shared underlying genetic aetiology. Current detection approaches rely on p-values of individual studies. However, differences in power between studies are not accounted for by p-values whereas Bayes’ factors (BFs) do, and may be approximated from summary statistics. We use simulation studies to compare the power of frequentist and Bayesian approaches to overlap analyses, and to decide on appropriate thresholds for comparison between the two methods. It is empirically illustrated that BFs have the advantage over p-values of a decreasing type I error rate as study size increases in single-disease associations. Consequently, the overlap analysis of traits from different-sized studies encounters issues in fair p-value threshold selection, while BFs are adjusted automatically. Extensive simulations show that Bayesian overlap analyses tend to have higher power than those that assess association strength with p-values, particularly in low power scenarios. An application of our methods is used to identify variants associated with both obesity and osteoarthritis.