About Phylogenetic Trees

Introduction

Phylogenetic trees are a method for determining evolutionary relationships between organisms, and require analysis of homologous sequences in different species. The premise behind phylogenetic analysis stems from an understanding of evolution and natural selection. Over time, species acquire mutations in their genomes, and advantageous mutations enable certain members of a species to pass on their genome (which contains these advantageous mutations), more easily than other members. During this process, species diverge from each other, and as mutations develop in the genomes of these diverged species, they become less closely related. Phylogenic trees track the history of species lineages as they change through time. Each species is related to any other by a common ancestor, or the last organism to which both modern species were related before branching occurred and caused new, diverged species.

For more information on Phylogenetic Trees, visit the Tree of Life Web Project .

The Maximum Parsimony Algorithm

Maximum Parsimony is a character-based algorithm that analyzes submitted DNA/RNA/protein sequences and finds a path between them that requires the fewest mutations. This is known as "economy of substitutions". The premise of this algorithm is that an evolutionary path between species is likely to occur in as few evolutionary steps as possible. The algorithm does not assume that the most parsimonious route is the exact route evolution follows; it only presumes that it is likely to occur. The challenge that the Maximum Parsimony algorithm attempts to resolve is creating all of the most parsimonious sequences, and choosing the tree that ultimately has the lowest score (lowest number of mutations).

The Maximum Parsimony method attempts to give to answers regarding phylogenetic trees, known as Small and Large Parsimony:

Small Parsimony: Given nucleotide sequences from different species in a pre-designed phylogenetic tree, the score of the tree (the number of mutations required to form the tree) is returned.

Large Parsimony: Given nucleotide sequences from different species, candidate trees are returned. However, if every single possible tree was created, millions of trees would be returned, and the program would not be very efficient. Therefore, the branch-and-bound method is employed, which calculates a tree, and then only continues to calculate trees that return a lower score. If a higher score occurs at any point, the tree is discarded in search of a better option.

To read the original paper on Maximum Parsimony, click here .