Friday, November 16, 2007

The Origins of Genome Architecture   posted by Razib @ 11/16/2007 10:43:00 AM

So picked up Michael Lynch's The Origins of Genome Architecture a bit worried, after all I don't have the marginal time to work through something like Genetics and Analysis of Quantitative Traits. Well, I didn't need to worry, I'm about half way through The Origins of Genome Architecture and it reads more like a manifesto than a text. Not that it isn't densely packed with references to the latest research (though I can see places where it is obvious it went to printing just a bit early), but the prose doesn't have the clunky flavor common to scientific textbooks coauthored by four or five different individuals. In other words, Lynch allows his voice to speak pretty clearly and unequivocally. Much of the material in The Origins of Genome Architecture is an amplification of previously published work (on some occasions he even simply interpolates prose you've seen in journal articles), which makes sense since Lynch notes that he's been working on this book for six years. Some of the articles he's published are obviously early leaks of ideas he formulated for book. In any case, I hope I won't ruin it for you when I tell you that Lynch makes the case that:

  • Population genetics can explain genomic architecture
  • Selection is less powerful across the arc of evolution than you would think
  • Population size matters a lot

A subheading for The Origins of Genome Architecture could have been "Why R.A. Fisher was wrong." Lynch's case hinges quite a bit (so far) on genetic draft and how it relates to effective population size.

The population genetics in this book is really low level, basically some algebraic manipulations which aren't difficult to follow if you simply read closely and think hard. He might refer to a diffusion equations, but Lynch won't derive it for you, you're presented with the neat formalism that pops out at the end of the process. Nevertheless, I do have to suggest that if you're someone without a biological education this book is going to be impenetrable. The population genetic math isn't very hairy, you can grasp the basic logic of the argument from the spare formalisms without the underlying nuts and bolts which connect the formulas, but the molecular genetic jargon will probably be too much for the typical uninitiated lay person. If your main exposure to biology is this weblog, then the stream of UTRs, spliceosomes, and poly(A) tails is really going to render much of the prose opaque and worthless. Though Lynch makes it clear that details are important, to keep the size of the book down he didn't elucidate what splicing in genetics means and why it's important. Much of the background information is assumed, so if you don't have that under your belt the exposition is worthless.