Lemme guess, you're also a fan of the Yankees, Red Wings, Duke basketball, and Florida football. You're kind of people sicken me.

This is consistent with the idea that trans variation segregates within populations, while cis variation differentiates species. I think G pointed this out in his post on the topic.

There are also pigmentation mutations in flies that have behavioral affects. I think either the ebony or tan mutations (or both) have marked effects on the behavior of Drosophila. Not surprisingly, some of the Sean Carrol-ites are looking into this.

it's a real bummer that the authors haven't responded to the comment

Recent genetic drift including founder effects would affect all genomic regions equally, but the candidate selected genes occur predominantly in genic regions, and preferentially include genes in functional classes that are plausible targets for recent adaptive changes. Selection is the only explanation consistent with all these features. 
 
I'm not an expert on the HapMap data, but aren't they enriched for SNPs in protein coding genes? Would this bias affect their results and conclusions?

Hold on a second there, champ. Should we vote for Kambiz or Shelley? You can't have it both ways, can you?

I think the article is here: http://arxiv.org/abs/0705.4062

I am very surprised at the diversity in sizes of organisms with similar phenotypic complexity. Anyone know what accounts for this? 
 
I'm assuming you mean "diversity in genome sizes". Here's a possible answer: http://mbe.oxfordjournals.org/cgi/content/abstract/23/2/450 
 
I saw this, but I wasn't going to blog it until the article comes out. The press release, as expected, provides very little information. For example, I can't tell how they measured mutations. Did they only count mutations in genes or is it all mutations? I have a feeling that they measured the number of mutations with phenotypic effects.

Nick Wade is a decoder.

One important distinction between correlation and regression that I don't think you explicitly state is that correlation describes the relationship between two independent variables, while regression (usually) describes the relationship between a dependent and an independent variable.

A few other people have hinted at it, but we can differentiated between neutral loss of function and selected lack of function. The standard popgen tests should be able to tell the difference. One simple prediction is that if loss of function is adaptive, we'd expect substitutions to be clustered in the region of the protein responsible for the functional differences (or the regulatory regions responsible for expression). If the loss of function was a neutral process, subsitutions should be scattered "randomly" throughout the gene.

Overall, he presents a sort of neutral theory of genome evolution, or at least the beginnings of one. 
 
Technically, it's a nearly neutral theory of genome evolution. And there are a few people getting tired of geneticists over-invoking population size to explain patterns in genetic data. The nearly-neutral theory has become the en vogue way to explain sequence polymorphism, divergence, and genome structure. 
 
[I tried posting a similar comment earlier, but it never made it through.]

that's what I think is the problem with this paper-- it assumes their test for selection has precisely the same power to detect selection in humans and in chimps 
 
Saying that's "the problem" with the paper neglects all the other problems. Not only does dn/ds suck balls as a method for detecting selection, but, when looking as such closely related species, how can you get by ignoring the intraspecific polymorphism? How many of the "fixed differences" in this paper are actually segregating as polymorphisms. 
 
Someone's gotta step up and do a big ole' MK test between humans and chimps. I'm not talking Bustamante et al, only polymorphism in humans, but something that incorporates chimp polymorphism.

Inversion polymorphisms are classic in Drosophila, but I don't know of any deletion polymorphisms from the classic literature. Deficiencies (large deletions) were used by geneticists, but they weren't natural occuring variation. They were created in the lab and used to map mutations. 
 
It appears that trying to cover you bases has backfired! 
 
And I'm surprised you didn't mention any of these studies.

The link to the paper isn't working, and I can't seem to find it on pubmed or on the Evolution website.

Another fucking genome scan, huh? Seriously, though, what are Andy Clark and Carlos Bustamante doing using anything other than the Applera resequencing data?

Touche, p-ter. Molly really fucked that one up. 
 
Someone needs to write a review of how a lot of recent findings in theoretical popgen show that we can't really detect selection using dna sequences because of a lack of power and/or too many false positives (ie, we're getting bad shit from both sides). I'm thinking either Przeworski or Andolfatto+Thornton.

I interpret a "genome scan" to mean a search of all loci in the genome to identify outliers (kind of like this). There are no a priori targets in a genome scan. Lahn's group and Reich's group both specifically targetted ASPM for resequencing, but determine significance using different approaches (simulation v. emperical distbn). That's why I don't consider these two papers "genome scans". But I could have a poor understanding of the terminology.

Two points: 
 
1. This isn't the type of genome scans Przeworski and crew were talking about, was it? Weren't they thinking more about the blind searchers for outlier loci? That's not what Lahn's peeps did, right? 
 
2. If you like ragging on genome scans, see this. But don't let Montse Aguade see you reading it.

But smaller populations will also have more slightly deleterious fixations, which could increase the rate of amino acid evolution. 
 
As for razib's little jab at the importance of gene expression and copy number: George Zhang's lab is one of the leaders in the field of gene duplication. I would hesitate to conclude that he's ignoring the role of duplication in evolution.