Species Tree Inference from Clades and Splits on Gene Trees

• Activity Number: 276
• Type: Topic Contributed
• Date/Time: Tuesday, August 5, 2014 : 8:30 AM to 10:20 AM
• Sponsor: International Indian Statistical Association
• Author(s): John Rhodes*+ and Elizabeth Allman and James Degnan
• Companies: University of Alaska Fairbanks and University of Alaska Fairbanks and University of New Mexico
• Keywords: species tree ; coalescent ; consensus tree

Abstract

One approach to the inference of species trees from multilocus data sets is to first infer gene trees by standard phylogenetic methods, and then use these to build some sort of a consensus tree. If the gene trees are assumed to differ primarily due to incomplete lineage sorting, then a good consensus method for this purpose should be one that is statistically consistent under the multispecies coalescent model. We first discuss how the STAR method of Liu et al. is such a method, and explain how STAR in fact uses only the summary statistics of clades frequencies on gene trees, and not the full topological gene trees frequencies. We also discuss alternative approaches, using clades or splits, and why these may be robust to certain types of gene tree inference error.