Mutation Rate Estimation and Classification of Driver and Passenger Genes in Cancer

• Activity Number: 512
• Type: Contributed
• Date/Time: Wednesday, August 4, 2010 : 10:30 AM to 12:20 PM
• Sponsor: Biometrics Section
• Author(s): Jie Ding*+ and Giovanni Parmigiani and Xiaogang Zhong
• Companies: Dana-Farber Cancer Institute and Harvard University and The Johns Hopkins University
• Keywords: Cancer mutation ; Driver gene ; Somatic mutation rate ; Markov chain Monte Carlo ; Bayesian analysis

Abstract

Genes with somatic mutations in cancer cells may be classified into two categories. One includes genes that are selected for by cancer (drivers). The other includes genes that are mutated by chance and don't contribute to the development of cancer (passengers). We propose Bayesian methods to estimate somatic mutation rates for genes and classify them into drivers and passengers using whole genome somatic mutation data set. Our methods can utilize different priors on mutation rates and incorporate both mutation type effects and sample effects. Markov chain Monte Carlo algorithms are designed to estimate mutation rates and other parameters of interest, such as the overall percentage of driver genes for a certain cancer type. Our methods are applied to both simulated and real data sets.