Re: What goes around comes around

From: Preston Garrison <>
Date: Fri Oct 07 2005 - 19:00:55 EDT


A few comments.

It seems that for you a class of evidence must be absolutely flawless
to count at all. There must be no anomalies, even if anomalies are
reasonably expected. You say, "It can't be a prediction if it is
explained away when it fails." Nonsense. A prediction that is born
out in 95% of cases is a pretty good prediction. The fact is that any
competent population geneticist could predict that things like
lineage sorting will happen if you look at enough cases. This is not
a "wild speculation." It's a reasonable prediction. In some cases you
can find out what happened by looking at the details, as someone
mentioned for a TE that was largely deleted. In some cases you can't.
That's genetics.

You dogmatically assert that no HERV can ever delete cleanly. First,
there's just no basis for such an absolute assertion in genetics. My
whole career in science could be marked off by the things that "can
never ever happen" that then happened. When I was in grad school it
was gospel, believe it or not, that antibodies do not bind to
denatured proteins. Surprise. They can! But all that aside, consider
this. Here's a sequence. It forms a tandem duplication, as often
happens. A HERV inserts in one copy. Later, homologous recombination
removes one copy of the duplication, with the HERV. Also a common
event. Appears to be a clean deletion. It's predictable that such
things will happen. These kind of events are not wild speculations.
Genomes are full of evidence for them, and watching cells in culture
will give you a lot more. It didn't even take a geneticist, only a
lowly biochemist to think of this one. So, part of what is sensibly
predicted is that there will be anomalies.

You say that you didn't mean to imply a high rate of homoplasies. You
said "...literally are many more of these mutations that cannot be
due to common descent" and "vast pool of repeated mutations." Sounds
like a high rate to me. But anyway. If you don't really have a high
rate of homoplasy, we're back to go. How do you account for the
original simple observation of thousands of TE insertions in the same
place in different species, quite apart from any phylogenetic
patterns. And remember, you don't just have to get an L1 or an Alu in
the position, it has to be the same subsubtype. It seems to me you
are brushing this off too easily, but I've given up on comprehending
your reasoning. If you have any other refs than the one from Pim, I
would appreciate them.

Your list of things that would convince you on evolution seem to
focus almost entirely on matters of mechanism. If matters of
mechanism are you're overriding concern, why not think about CD + ID
as a solution?

One more thing. Why would making a cell in vitro affect your
judgement of evolution?


Received on Fri Oct 7 19:02:50 2005

This archive was generated by hypermail 2.1.8 : Fri Oct 07 2005 - 19:02:50 EDT