I wonder if you could do something with Monte Carlo simulation?--
either ask the question "How often will one tree produce sequences this
unrelated?" or "How often will multiple disjoint trees produce sequences
this related?" You will need a reasonably plausible model of sequence
evolution, which may not be easy to come by: but without one I'm not
sure *any* statistical test will do much for you.
Mary Kuhner mkkuhner at genetics.washington.edu
I do not receive posts from the following systems because they tolerate
abuse of Usenet: interramp.com psi.com scruz.net
If you wish me to see your message anyway please use email.