PgmNr P2148: A maximum pseudo-likelihood approach for estimating species trees.

Authors:
L. liu ¹ ; L. Yu ² ; S. Edwards ³

---

Institutes
1) University of Georgia, Athens, GE; 2) Georgia Southern University, Statesboro, GA; 3) Harvard University, Cambridge, MA.

Abstract:

Several phylogenetic approaches have been developed to estimate species trees from collections of gene trees. We propose to use a pseudo-likelihood function of the species tree to obtain Maximum Pseudo-likelihood Estimates of Species Trees (MP-EST), with branch lengths of the species tree in coalescent units. The MP-EST method estimates species trees from a set of gene trees by maximizing a pseudo-likelihood function. The program can run independent searches (chains) in parallel using the parallel version of the program. Each chain starts with a different seed. The program will find the estimate of the species tree with the largest pseudo-likelihood score across chains.