Setting the boundaries of a listing such as this is a difficult task, but there have to be some boundaries. I have decided that this list will not include:

• Data conversion utilities. These are important and useful programs that convert your data from the format needed by one program to the format needed by another. As they do not themselves perform any data analysis, I have reluctantly decided not to list them. One, Tonex was formerly listed here and has been removed from the list.
• Clustering programs that are not intended to infer phylogenies. There is a large and active literature on clustering methods, with many programs in distribution. However making clusters and inferring phylogenies are not the same thing (though closely related). I have tried to decide which packages had some material on inferring phylogenies, and include only those.
• Tree alignment programs that do not infer trees. Programs or servers that use tree-based alignment of sequences, but do not allow you to output the resulting tree, fall slightly outside of this listing, as any tree that they have is not visible to the outside world. This means that some ClustalW servers are not listed here (those that do not allow you access to the tree).
• Programs for detecting evidence of recombination that are not using trees intensively enough. For example, there are useful and helpful tests that look at pairs of sites but not at anything involving trees more directly. This is a particularly difficult boundary to define but there has to be one here somewhere.
• Sequence editors that do not submit jobs to phylogeny programs. I have listed a number of sequence-editing environments that allow one to automatically invoke phylogeny programs on selected sets of sequences or sets of sites. Sequence editors that do not have these job submission capabilities lie outside the domain of this listing.
• Coalescent programs that do not infer trees. This is a tough one. There are now a number of programs that add up likelihoods or Bayesian posteriors over different possible coalescent trees of genes within species. Our own lab is distributing one such package. This category of programs is growing rapidly. Many of them infer population parameters or genetic parameters, or times of most recent common ancestry of mutants. Wonderful and important stuff, but as it does not focus on between-species phylogeny, I am going to exclude most of these programs. The exceptions will be those that report inferences of the coalescent tree itself, or those that infer the between-species phylogeny. This is going to be a hard boundary to define as coalescent programs proliferate.

I will continue to agonize over these boundaries, but for now this is the policy. If you feel that these policies are wrong, you can always (a) argue with me, or (b) make your own phylogeny programs web pages.