|
Abstract:
|
The advent of microarray technology has enabled biomedical researchers to monitor changes in the expression levels of thousands of genes. Until recently, however, the mechanisms driving these changes were harder to study in a similarly high-throughput level. A recent technological innovation, chromatin immuno-precipitation coupled with microarray (chip) analysis (hence the name ChIP-chip), now makes it possible for researchers to identify regions of a given genome bound by specific DNA binding proteins. During this talk, I will present a flexible and robust Bayesian hierarchical model for the analysis of ChIP-chip data that incorporates the spatial structure (probe dependence) while borrowing strength from all the probes. Detection of bound regions will be based on the posterior distribution of the parameters. The model will be illustrated using Affymetrix tiling arrays.
|
