Subject: Re: chapter 13 in Hartl & Jones
Date: Thu, 16 Apr 1998 09:55:19 -0400
From: dhartl@oeb.harvard.edu (Daniel Hartl)
To: "Ussery, David " <ussery@roanoke.edu>

Dear Dave:

Thanks for your message. I hope I can clear it up. That the EXPECTED ratio
of transitions : transversions (assuming complete randomness) is 1 : 2 is
clear from enumeration, since any base, when it mutates, can undergo a
transition in only one way (e. g. A -> G) but a transversion in two ways
(e. g., A -> T and A -> C).

The DATA come from molecular evolutionary studies of pseudogenes and other unconstrained sequences, in which the ratios of various types of
substitutions observed among species should be equal to the ratios of the
mutation rates. The obesrved ratio of transitions : transversions is 2 : 1
(approximately).

One reference is M. Nei, Molecular Evolutionary Genetics, p. 28:

"The relative mutation rates obtained in this way [the way I described] are
presented in Table 3.4. It is clear that the frequency (59.3%) of
transitional changes is much higher than the expected (33.3%)."

Other books and papers in molecular evolution give the same story. In the
real world the changes are the reverse of those expected w/randomness.

I suspect the main reason is that most types of chemical modifications to
the bases (deamination, demethylation, etc.) are likely to cause mispairing
of a type leading to a transition. Then, too, maybe mismatch repair is more
efficient in purine-purine or pyrimidine-pyrimidine mispairs, which would
also favor transitions.

I hope this helps. If it does not, please let me know.

I do not know why this is not stated in other textbooks. To me it seems
like an interesting and important observation, and also unexpected.

Best regards to you and your class.

Now I'm off to give MY lecture!

Dan

>Thursday, 16 April, 1998; 9:13 EST
>
>Dear Dan,
>
> Yesterday in our Genetics, we covered chapter 13 - on page 557 of
>the text, it states that the ratio of transitions to transversions
>should be expected to be 1:2 - although you then say that in fact the
>ratio is 2:1, and then talk about bias and the genetic code. First, is
>this really true that the ratio is opposite of what's expected, (I found
>the wording a bit confusing) and if so, is there any known basis for
>this observed difference? I (simplistically) thought perhaps it was a
>typo, but my genetics class read it differently - they all thought it
>was NOT a typo. (Actually I'm assuming I probably read it wrong and
>hope you'll come to the rescue of my class!).
>
>Here's the URL for yesterday's lecture notes:
>http://www2.roanoke.edu/biology/dussery/Genetics980415f.htm
>
> I've tried to find something about this in two other genetics texts
>- Griffiths of course has a nice dicussion about
>transitions/transversion, but says nothing about the ratio. Klug &
>Cummings doesn't appear to really discuss this at all (I couldn't find
>"transition" or "transversion" in their index, nor did I see a Chapter
>that dealt mainly with molecular mechanisms of mutation).
>
>
>with kind regards,
>
>Dave Ussery