Regarding David Maddison's question about convergent evolution and how
to determine if genes are part of the same tree or not:
R. Doolittle had a review in TIBS last year about such issues.
In addition there is some published work relating to this question from
people working on in-vitro directed evolution. In these studies,
regions of genes or entire genes can either be replced by random
sequences or mutagenized at very high rates. These genes can then be
selected for specific functions (such as catalytic RNA activity or DNA
binding activity of the encoded protein). Thus these systems can be
used to determine how easy it is to converge on the same sequence by
strong selection. Specifically Larry Loeb and colleagues at U.
Washington have done some work relating to this (e.g., see Gene
137:41-47). Not all of the people doing such work have addressed the
issue of convergence but there is probably a lot of information in their
studies that might be of relevance.
Jonathan A. Eisen
Department of Biological Sciences
Stanford, CA 94305-5020
jeisen at leland.stanford.edu