Bayesian Methods in Genomics: Searching for Unity in Diverse Data Sources

• Activity Number: 543
• Type: Roundtables
• Date/Time: Wednesday, August 4, 2010 : 12:30 PM to 1:50 PM
• Sponsor: Section on Bayesian Statistical Science
• Author(s): Debashis Ghosh*+ and Bhramar Mukherjee*+
• Companies: Pennsylvania State University and University of Michigan
• Keywords: Genomics ; Personalized Medicine ; Prediction Models ; Gibbs Sampling

Abstract

With the completion of the Human Genome Project, the international HapMap project and the 1000 Genomes project, statisticians, geneticists and bioinformaticians are faced with an enormous challenge of integrating diverse data sources in researching disease etiology, prognosis and treatment. The genetic architecture of a disease is then analyzed in conjunction with exposure history to behavioral and environmental factors. In this complex modeling era, Bayesian methods and principles are being invoked for data imputation, meta analysis, variable selection, eliciting priors from functional databases, network and pathway modeling and in many other contexts. In this round table discussion we will summarize a few recent important Bayesian applications in this area and identify emerging areas where involvement of Bayesian researchers may be beneficial.