|
Activity Number:
|
225
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 3, 2009 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Bayesian Statistical Science
|
| Abstract - #305864 |
|
Title:
|
Impact of Informative Prior in Discrete Bayesian Graphical Models: Application to Genome-Wide Association Studies
|
|
Author(s):
|
Jinnan Liu*+ and Laurent Briollais and Adrian Dobra and Helene Massam
|
|
Companies:
|
Samuel Lunenfeld Research Institute of Mount Sinai Hospital and Samuel Lunenfeld Research Institute of Mount Sinai Hospital and University of Washington and York University
|
|
Address:
|
, , ,
|
|
Keywords:
|
Contingency tables ; Conjugate priors ; stochastic search ; cancer susceptibility genes
|
|
Abstract:
|
High-throughput sequencing studies together with clinical and physiological data produce large amounts of biological information that is used for molecular phenotyping of many diseases. Combined with efficient algorithms that can explore the model space rapidly, Bayesian graphical models allow an efficient evaluation of a large number of multi-marker models, where each model embodies complex interactions among the markers. Our goal is to use this framework to perform genome-wide association studies and identify important combinations of biomarkers that will improve the prediction of breast cancer risk. After deriving the analytical properties of discrete Bayesian graphical models, we will assess the importance of expert biological knowledge for new genes discovery.
|
