In article <610o18$mgl at net.bio.net>,
John Clement <clement at mail.utexas.edu> wrote:
> I was wondering if there are methods available for determining
> if there are branches on an evolutionary tree where amino acid
> sequences have undergone signficant amounts of non-conservative
> change in their side-chain chemistry.
One can first look for a difference in the ratio of
synonymous to nonsynonymous substitutions. See AIDS Res
Human Retroviruses 12(18): 1681-1686 (1996) for an example.
This will tell you if different branches have differing
rates of amino acid change, while having the same rate
of mutation. Be careful though, as synonymous sites can
become saturated. You need to use a model which takes into
account the fact that as the substitutins get closer to
saturation, homoplasy (multiple hits at the same site)
increases.
You need to do this, because of course each branch
of an evolutionary tree has a different length. If we
assume a constant, reliable molecular clock, we say that the
different lengths represent different fimes of divergence
from a common ancestor. If we want to show that the clock
ticks faster on one branch than another, we need both
fossil evidence of the time of divergence (or in the case
of HIV-1 we can use frozen blood samples and other old
molecular data) and accurate distance measurments.
If we want to show that two branches mutate at
roughly the same rate (synonymous mutations rates are equal),
yet one branch has different selection pressure (resulting in a
difference in rate of nonsynonymous change) we measure the
syn/nonsyn ratio, or the difference between syn and nonsyn
on each branch.
After doing this, you can measure the nonsynonymous
distances (using a method which will treat all nonsynonymous
mutations equally) and measure the protein sequence distances
with a PAM250 or other matrix which weights different
substitutions differently, and see if there is a difference.
i.e. do two lineages with equal rates of protein change
overall show different rates of conservative change?
The PIMA program package {Protein Engineering 5(1):
35-41 (1992)} can use a stuctural matrix {Henikoff and
Henikoff, Proteins: Structure Function and Genetics 17:
49-61 (1993} to compute the distances by weighting amino
acid changes by side chain structure.
You might want to be careful about what window size
you use. I know of some proteins in some lineages of HIV-1
which evolve in pretty much the same pattern overall, but
when you look at different functional regions, one lineage
tends to be variable in one region, while another is variable
in another region. i.e. two proteins can keep the same level
of overall "fitness" by two different stategies, one might
increase the binding affinity while allowing catalytic rate to
decrease a bit, the other might increase catalytic rate but
allow binding affinity to decrease, and third might decrease
both binding and catalytic rate, but have mutations in the
non-coding promoter region to increase the transcription and
thus concentration in the cell.
--
____________________________________________________________________
|Brian T. Foley btf at t10.lanl.gov |
|HIV Database (505) 665-1970 |
|Los Alamos National Lab http://hiv-web.lanl.gov/index.html |
|Los Alamos, NM 87544 U.S.A. http://www.t10.lanl.gov/~btf/home.html |
|____________________________________________________________________|
