Integrative analysis of genetic association data from molecular and complex traits has emerged as a practical approach to investigate disease etiology. In this talk, we will focus on two prevailing types of integrative analysis methods: co-localization analysis and transcriptome-wise association study (TWAS). The former method attempts to assess the overlappings of causal genetic variants from different types of traits, and the latter approach aims to identify causal molecular phenotypes leading to complex traits. A critical challenge for both kinds of integrative analysis lies in the effective characterization and incorporation of genetic association evidence and the corresponding uncertainty. By taking advantage of probabilistic summaries from Bayesian multi-variant genetic association analysis, we demonstrate a set of rigorous and intuitive statistical solutions to both types of integrative analyses. Finally, we discuss the promise and limitations of the current integrative analysis, as well as the statistical challenges for future work.
The Statistics Seminars for Winter Quarter will be held in Room 380C of Sloan Mathematics Center in the Main Quad at 4:30pm on Tuesdays. Refreshments are served at 4pm in the Lounge on the first floor of Sequoia Hall.