ACCURACY OF PHYLOGENETIC-ESTIMATION METHODS UNDER UNEQUAL EVOLUTIONARY RATES.

A comparative study of the accuracy of three different approaches to phylogenetic estimation was made on simulated data with differing rates of change in different lineages. The three approaches were maximum likelihood, maximum parsimony, and phenetic clustering. The data were generated by simulating genetic drift with different population sizes over a simple four-species tree topology. Although the accuracy of all three approaches was found to be dependent on the number of loci (characters), maximum likelihood was found to perform considerably and consistently better than maximum parsimony or phenetic clustering.