Bayesian Inference of Evolutionary History of Populations/Species Based on Importance Sampling of Gene Trees

• Activity Number: 354
• Type: Contributed
• Date/Time: Tuesday, August 11, 2015 : 10:30 AM to 12:20 PM
• Sponsor: Section on Bayesian Statistical Science
• Author(s): Yujin Chung*
• Keywords: Bayesian approach ; Importance sampling ; coalescent process ; Isolation with migration model ; MCMC ; Evolution

Abstract

Associated the advancing technologies of DNA sequencing, the study of evolution catalyzes interdisciplinary research that uses evolutionary knowledge or statistical methods in evolution enabling us to better understand genetic disorders. Estimating the demographic history of a group of populations/species from genetic data has been one of major interests in evolution. We propose a new Bayesian model to infer the demographic history of populations/species in the presence of migrations between populations from genetic data. The new method does not suffer from mixing problems that existing methods commonly have. Through MCMC, we first simulate coalescent trees carrying the information when and which lineages coalesced backward in time but not migration traces. Using the simulated genealogies, the posterior density of all the demographic parameters is approximated. This can be done without introducing biases in estimations by applying importance sampling and by exactly computing the conditional probability of a coalescent tree. In virtue of importance sampling, we can use the same simulated coalescent trees to infer different demographic models. A simulation study will be presented.