|
Activity Number:
|
310
- SPEED:Statistical Methods for GWAs, Genetics, Genomics, and Other Omics Studies, Part 2
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Tuesday, July 30, 2019 : 9:25 AM to 10:10 AM
|
|
Sponsor:
|
Biometrics Section
|
|
Abstract #307696
|
|
|
Title:
|
A Hierarchical Pitman-Yor Model for the Evolution of Phenotype Distribution on a Phylogenetic Tree
|
|
Author(s):
|
Hanxi Sun* and Heejung Shim and Vinayak Rao
|
|
Companies:
|
Purdue Statistics and University of Melbourne, Australia and Purdue University
|
|
Keywords:
|
Bayesian Nonparametric;
Phylogenetics;
Particle MCMC
|
|
Abstract:
|
Ecological evidence suggests the existence of rapid evolution periods ("jumps") in phylogenetic trees. Locating such jumps is crucial towards understanding observed phenotypic diversity. Recently, Ansari and Didelot (2016) introduced hierarchical Dirichlet processes to model the evolution of phenotype distributions. They, however, ignored dependency across the tree for computational reasons. In this work, we propose a new Bayesian framework that employs hierarchical Pitman-Yor processes (HPYPs). We exploit the marginalization property of PYPs to develop a tractable algorithm that still enables sharing between different nodes in the tree. The proposed method also bridges the gap between traditional random walk approaches and pure jump models of evolution. We show the effectiveness of the algorithm on both synthetic and real datasets.
Ansari, M. A., & Didelot, X. (2016). Bayesian inference of the evolution of a phenotype distribution on a phylogenetic tree. Genetics.
|
Authors who are presenting talks have a * after their name.
Back to the full JSM 2019 program
|
