Wednesday, November 08, 2006

Neanderthal introgression & microcephalin   posted by gcochran @ 11/08/2006 05:53:00 AM
Share/Bookmark

A few years ago, I was thinking about Out-of-Africa, and it occurred to me that we probably picked up a lot of favorable alleles from Neanderthals, the logic being that such alleles would have probability 2s per copy (Haldane) of reaching high frequency (maybe even going to fixation). The chances of acquisition of a neutral allele (from a single introduced copy) was 1/2Ne; so, for s = 5% and an effective population of 10,000, that allele with the 5% advantage was 2000 times more likely to make the interspecies jump. Therefore mtDNA stats really told you nothing.

So as we spread out into Neanderthal territory, even a few tens of interspecies matings would have let us scarf up most of their good genes. Individuals from every pair of mammalian sister species with comparably recent common ancestry are interfertile, so it was a good bet that this actually happened. This seems to happen frequently in invasive weed species, by the way. They arrive, they hang around the docks for a few years, stealing the good alleles from related local species - then they go forth and conquer.

Looking at the new article in PNAS, I'd say that Bruce Lahn and company have probably found one. Read it.

Now if the Neanderthals were really effectively isolated before we expanded into their territory, they'd have a lot of significantly different alleles. Some would have involved various kinds of regional adaptations, which might be a good thing to have in Eurasia. (MC1R?) But it's entirely possible that some alleles solved adaptive problems that had existed in Africa as well - but solved them better.

Brains had expanded over the last half-million years in both Africans and Neanderthals, but it seems likely that those changes in size and structure were driven by different mutations, just as light skin in Europe and East Asia was. The Neanderthals had slightly bigger brains than Africans: obviously those brains were useful for _something_. Anyhow, in this kind of convergent evolution of sister species, there can be lots of alleles worth stealing. When we select for the same trait in multiple lines, sometimes we get higher values of that trait by hybridizing a couple of the best-performing lines. Also, since the Neanderthals were ecologically different (cold weather, high risk hunters & pure carnivores), they might have been able to evolve some adaptations that just couldn't happen in Africa (different constraints, different topography of the fitness surface).

Maybe Africans and Neanderthals 'nicked'. Any farm boy, looking at the timeline of African expansion, encounter with Neanderthals, and the subsequent 'great leap forward', should have suspected hybrid vigor. It's corny but it makes sense.

So when you think about the cultural explosion that occurred shortly after we overwhelmed the Neanderthals (cave paintings, sculptures, new tools and weapons, all that jazz) - well, you have to wonder if assimilating a passel of adaptive alleles in a few thousand years, way more than the typical number that would arise and become established over such a short time span, didn't give us a hell of a boost. There are signs of behavioral modernity a bit earlier in Africa - but those ostrich eggshells are dull as hell compared to Gravettian cave paintings. Expansion out of Africa must itself be a sign of new capabilities (I'd bet on sophisticated language) but you only see full-fledged behavioral modernity in the European Upper Paleolithic... Judging from neutral genes, it can't have happened often, but those few furtive human-Neanderthal couplings may well played a crucial role in the future development of the human race. I'm sure that this notion will suggest new pick-up lines to some readers.

If this pans out the way we think it will, introgression from Neanderthals (and maybe with other archaics) may have been one of the two fundamental patterns underlying recent human evolution.

Update by Darth Quixote: John Hawks has more here and here.