Re: Doogie Mice

Robert J. Bradbury (
Thu, 2 Sep 1999 19:24:46 -0700 (PDT)

On Thu, 2 Sep 1999 wrote:

> In a message dated 9/2/99 14:03:01, writes:

> >That requires very accurate gene duplication of the gene in
> >question (without duplicating any negative genes), or a very
> >specific mutation in the NMDA expression or transcription
> >regulatory region(s).
> Mutations to regulatory regions do not need to be specific at all.
> As a rule, most mutations in regulatory regions will either
> increase or decrease expression.

I don't think we disagree here. What I was trying to say was that they have to be *specific* for increases in expression. Changing a GC to a CG may increase expression, while changing a GC to an AT may decrease expression.

> Increases aren't particularly hard to come by either; although
> most mutation will decrease the effect of the region,

yes. The number of mutations away from the optimal situation is probably much higher than the number of mutations that improve on an already "good" situation.

> many regulators are repressors. Most regulatory regions are
> several bases off consensus (= most active, presumably)
> so *increases* to regulatory activity aren't that hard
> to come by anyway.

I think we would have to discuss this more carefully. While in bacterial genomes I might agree that repressors tend to be more common, I suspect that in higher organisms, promoters would tend to be the dominant factors. I suspect we need more information to know this for certain.

> Variation in regulatory regions is turning out to be quite common
> and seems to underly most natural quantitative genetic variation.
I wouldn't say "most", since statistically the genes are *much* larger than the regulatory regions, so there will be more variation in the genes than in the regulatory regions. Now whether this balances out in terms of mutations in gene "critical regions" being roughly equivalent to mutations in the "regulatory regions" (in terms of a rough equivalence in the number of bases) I suspect is an open question.

I would say that *most* of the quantitative genetic variation that I have seen to date involves mutations in the gene coding regions. I would allow that there are selection biases that may be creating that impression, so we would need a lot more work to determine whether that is the reality.