No doubt you've seen Lewin's report on molecular drive (Science 218 552-3). At least he's got right that molecular
drive is a system of (a) fixation and (b) with a particularly synchronous pattern of fixation. However, there are a few remarks
by commentators which puzzle me, at least those of Jeffreys. We've replied to these and I'm enclosing a copy of our
letter to Science. With respect to Alu I guess he's picked up on your remarks at Heidelberg i.e. the absence of a difference
in within- and between-species divergence in the primate Alu. Where are these data published? I'm very interested to look
at them. I'm not sure if they were your own data. If they are, could you send me a copy of any paper you have on this,
if it's not yet published?
You'll see from our enclosed letter, that we regard the observed levels of divergence at any one time to reflect the differences
in rates between homogenization and mutation. There are many factors influencing this, not least the family itself. What
really interests us is if there has ever been a systematic search for a species diagnostic variant, indicative of separate
homogenization in each primate species? Unless this is done, in the way for example Steve Brown did in the MIF-family, then
it's always difficult to interpret divergence levels. This is especially difficult if the levels are taken from a handful
of clones in each species. A family like Alu with 500,000 members is bound to have both divergent isolated members and also
a subfamily structure. The question is what do the clones really represent? We know that MIF in 5 species of rodents has
several subfamilies (i.e. partial homogenization of a diagnostic variant); however surprisingly these subfamilies are present
in all species - but to different extents (data in preparation). There are several complex interpretations of these patterns,
but nevertheless the differences in abundance of the different subfamilies between species, suggests that MIF is undergoing
a process of homogenization in each species and that different subfamilies are either on the way in or on the way out. Or
alternatively the constraints on large family total homogenization might lead to an equilibrium situation.
All these types of data come from whole-family analysis. They would not have been revealed by sequencing a few clones - although
we have this as well in MIF which tell us other things as well - but that's another story. So even though there has been
a clear homogenization between rodent and human Alu family counterparts (see enclosed) - revealed mainly by whole family studies
- the question is, is it really as slow as to be only detectable when one goes as far back in the species phylogeny as rodent
versus man. Could it be that there are also distinct whole family or subfamily differences between primate species indicative
of a faster rate of homogenization?
Is there anything in the Alu data that throws any light on this? Or better, are there any data that definitely rules out some
homogenization since primate species separation? Given the debate now initiated in Science, we are very eager to know.