Lecture 18 Determining Phylogenies by Distance Matrix February 29 , 1996 Lecturer : Joe Felsenstein

The sites that arouse skepticism in the parsimony method are the ones with the greatest number of changes. Compatibility is a technique related to parsimony, and is well suited to trees for which parsimony would produce frequent changes between states. Unfortunately, however, compatibility also treats small noise artifacts as significant. Definition 18.1: A site is compatible with a tree if and only if each state of that site arises no more than once on the tree, i.e., the number of changes for that site is one less than the number of states of that site. Compatibility attempts to maximize the number of such sites. Definition 18.2: Two sites are compatible if and only if there is a tree T such that each site is compatible with T. How can compatibility be used to construct a phylogenetic tree? Suppose only two states are considered as possibilities (i.e. f0; 1g or fY;Rg, where Y represents C or T , and R represents A or G). For two sites, count how many of the four pairs of states occur together in any species: the two sites are compatible if and only if fewer than four pairs occur together. Example 18.3: In the four species labeling the leaves below, only three of the four pairs of states occur together, so the two sites are compatible. The tree itself certifies that compatibility. 00 01 11 11 \ / \ / 01 11 \ / \ / 01 87