Links between maximum likelihood and maximum parsimony under a simple model of site substitution (Q1363041)

This paper deals with the problem of inferring phylogenetic trees from genetic sequence data. Two standard methods of tree reconstruction are compared to each other: the maximum likelihood approach maximizing a certain likelihood function for the observed data set as well as the maximum parsimony method minimizing the total number of mutations required to explain the data from present day species. Under the assumptions of a simple nucleotide substitution process called fully symmetric model, it is shown that maximum likelihood and maximum parsimony are equivalent in the selection of phylogenetic trees and in the reconstruction of ancestral states on any given tree. Some of these results generalize earlier results from two to an arbitrary number of states.