Semiparametric Bayesian Analysis of High-Throughput Array CGH Data

• Activity Number: 242
• Type: Contributed
• Date/Time: Monday, August 2, 2010 : 2:00 PM to 3:50 PM
• Sponsor: Section on Bayesian Statistical Science
• Author(s): Subharup Guha*+
• Companies: University of Missouri
• Keywords: Generalized Polya urn process ; Copy number ; Genomic alterations ; MCMC ; Tumor ; Lung cancer

Abstract

Array CGH has emerged as a powerful technique for identifying genomic alterations potentially linked to various kinds of cancer. Adopting a semiparametric Bayesian approach, we model the dependence among neighboring copy number alterations using a first-order infinite-dimensional generalized Polya urn process. The model overcomes several shortcomings of existing approaches based on the hidden Markov model, while accounting for multi-modality and skewness.

For high-throughput array CGH data, current MCMC techniques take an order of magnitude longer than alternative approaches that are often less accurate. We apply a cost-effective MCMC strategy to perform a fully Bayesian analysis of high-resolution data on lung cancer. The results demonstrate the ability of the proposed approach to easily detect large-scale trends as well as localized changes in copy number.