« The past is not the present.... | Gene Expression Front Page | More babies for South Koreans? »
April 02, 2004

What's Sex got to do with it?

As some of you may have noticed, I really stepped in it in the comments section over at The Panda's Thumb during their first week in operation. It was an innocent mistake on my part, I assure you, though everything worked out ok for me in the end, because it forced them to take a hard look at a couple of posts I made to their site.

Anyway, to recapitulate a little: I was explaining how, when I first became interested in molecular biology a few years back, that one of the questions that got me started was calculating the odds that chance and time alone could explain the complexity of the human genome. Based on reports coming in, there were a couple of billions of base pairs in the human genome. Given there were four letters in the genetic alphabet, I reasoned there were 4 raised to the power of 2 billion possibilities to choose among ? which is a truly mind-boggling, trans-astronomical number.

I then re-traced for them the thought process by which I got that number down:

First, there is all of the junk, of course: and then second, the "step-wise" process of selection, by which first one mutation takes place and is fixed, before the next one begins.

This is stuff you find in every biology text that deals with the question, and went a long way, though not all the way, towards answering my question.

Next I found an amateur site on the web somewhere, in which a guy actually calculates the number of mutations that could have been fixed in evolutionary time, based on some reasonable assumptions about average population size and average number of generations, which he got from Fisher I think. The number he arrives at is 135, which, he points out, is roughly the number of mutations fixed on a typical protein.

That got the problem much closer to solution, but still not all the way. It was only when I realized -- operating completely on my own at this point, because I couldn?t find anything in the literature -- that in a large, interbreeding population every protein in a species is undergoing simultaneous and, as it were, parallel evolution, that the problem was solved to my own satisfaction. The number 135 was indeed the magic number.

Anyway, since then I?ve been thinking about something related: sex is necessary for the process to work,. Or, more specifically, sex plus sexual cross-over between homologous regions in each chromosome pair, is required before there can be massively parallel evolution in the genome. Surely this is not an original insight with me, and yet I still see articles published in Nature and elsewhere from time to time speculating on the evolutionary function of sex.

Does anybody know anything about this in the literature, or am I (much more probably) guilty of some fundamental misunderstanding here?

Posted by lukelea at 07:21 AM