CVD forum talk: A Bayesian approach to the overlap analysis of epidemiologically linked traits

Add to your list(s) Download to your calendar using vCal

• Dr Jennifer Asimit, Wellcome Trust Sanger Institute
• Wednesday 27 May 2015, 14:00-15:00
• Thomas Strangeways Room, Strangeways Research Laboratory.

If you have a question about this talk, please contact dff21.

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.

This talk is part of the Cambridge Cardiovascular Seminar Series series.

This talk is included in these lists:

• Cambridge Cardiovascular Seminar Series
• Cambridge Centre for Data-Driven Discovery (C2D3)
• Cambridge talks
• Cardiovascular Epidemiology Unit Special Seminars
• Chris Davis' list
• Interested Talks
• Thomas Strangeways Room, Strangeways Research Laboratory
• Trust & Technology Initiative - interesting events
• bld31
• my_list
• ndk22's list
• ob366-ai4er
• rp587

Note that ex-directory lists are not shown.