Identifying Genes Under Selection Using Generalized Linear Mixed Models

• Activity Number: 301
• Type: Contributed
• Date/Time: Tuesday, August 3, 2010 : 8:30 AM to 10:20 AM
• Sponsor: IMS
• Author(s): Kirsten Elise Eilertson*+ and Jim Booth and Carlos Bustamante
• Companies: Cornell University and Cornell University and Stanford University
• Keywords: Generalized linear mixed models ; genetics ; natural selection ; Laplace approximation

Abstract

We present an approach for identifying genes under selection using polymorphism and divergence data from synonymous and nonsynonymous sites within genes. A generalized linear mixed model is used to model the relationship between the number of polymorphic/divergent mutations and synonymous/nonsynonymous sites. Based on the theory behind the McDonald-Kreitman statistic, we use the estimated fixed and random effects of the model to identify genes under selection. The model is fit in both the standard and Bayesian settings. The proposed methodology is designed to analyze several thousand genes from the same phylogeny. Using simulated data we compare our method to existing methods for detecting genes under selection, the MK statistic, and MKprf. The simulations show the GLMM method does an excellent job of identifying genes under selection while maintaining a low false positive rate.