From: Preston Garrison (garrisonp@uthscsa.edu)
Date: Sun Apr 13 2003 - 04:36:27 EDT
>
>Josh: The following quote is important for my general argument:
>
>"Although HERVs are known to be preferentially integrated into GC-
>and Alu- rich actively transcribed and early replicated areas of
>human chromosomes, the patterns of various HERVs and their LTRs
>within the human genome are diverse. For some, there is only slight
>unevenness in the distribution over chromosomes... IN contrast
>10,000-25,000 copies of HERV-K related LTRs are highly
>over-represented on human chromosomes 3 and 16, and underrepresented
>on chromosomes 8, 11, 13, and X." (Bioessays 22:161-171).
>
>There seems to be more to this than complete random access to the genome.
>
>Josh
>
Josh,
I don't think you can make your argument work. Viral and transposon
insertions are not completely random - you are right about that - but
they are nowhere near specific enough to account for all cross
species insertions at exactly the same site and orientation - often
in multiple species. Remember, too, that the common descent
"hypothesis" predicts phylogenetically coherent patterns for a given
insertion. An insertion should be present in a set of species that
makes sense for the insertion to have occurred at one point in the
phylogenetic tree. I think that is almost always the case in the
examples studied, but there may be rare anomolies from deletion of an
element. One also expects that the age of insertions determined from
their phylogenetic pattern should match their age estimated from
sequence divergence from the consensus for their class. Again, I
think this is generally true, with some anomolies from things like
gene conversion.
The elements we are discussing can be regarded as one form of a more
general category, which is mutations that are complex enough that
they are very unlikely to have occurred twice in exactly the same
way. Another example is chromosomal rearrangements - inversions,
translocations, duplicative transpositions, etc. The sequence
junctions in a rearrangement are analogous to mobile element
insertion sites in this regard. (There is the caveat that some
rearrangements depend on homologous recombination (often between
different copies of a mobile element like L1), and so they can recur.
The Philadelphia chromosome in myelogenous leukemia is an example.)
If common descent is true, it should be possible to work out a
coherent history of, for instance, the rearrangements of the
mammalian genome during the evolution of mammals. In fact, that
project is well under way. I haven't followed it very closely, as it
is rather complicated, but I think it is proving to be quite
workable. The syntenic regions (containing the same genes) of mouse
and human chromosomes have, of course been worked out in great
detail. Assuming that the history of rearrangements can be worked out
in a coherent scheme that matches the expected phylogenetic tree
without undo complexity, that seems like an awfully strong evidence
for common descent.
Preston G.
-- Preston Garrison, Ph.D. Instructor UTHSCSA Biochem. Dept. MSC 7760 Insert the usual disclaimers here. 7703 Floyd Curl Dr. San Antonio, TX 78229-3900 garrisonp@uthscsa.edu 210-567-3702 http://biochem.uthscsa.edu/~barnes
This archive was generated by hypermail 2.1.4 : Sun Apr 13 2003 - 04:38:02 EDT