Dr. Sylvia Richardson of the University of Cambridge delivers her featured lecture as the recipient of the 2023 Michael Boehnke Distinguished Visiting Scholar Award from the Department of Biostatistics at the University of Michigan.

Date: Wednesday, September 27
Location: Rackham Graduate School Amphitheatre

TITLE: Scalable Bayesian inference for high dimensional genomics

ABSTRACT: Through biotechnological advances, we are able to probe deeply into biological mechanisms and collect large quantities of molecular data. The new scientific questions which arise alongside create stimulating challenges to Bayesian style analyses of such data.

In this talk, I will discuss two pervasive high dimensional tasks in integrative genomics: (i) uncovering multivariate associations between large sets of predictors (genetic variants) and large set of responses (gene or protein expression), with particular interest in finding predictors associated with multiple responses (e.g. hotspots in eQTL studies); and (ii) performing scalable inference of conditional independent networks comprising large sets of features, and joint inference of related networks. In both cases, I will outline the formulation of suitable hierarchical structures which allow borrowing of information, briefly discuss the computational strategies that were developed to enable scalability of Bayesian inference to a large number of features and illustrate the performance of the models on genomics case studies.