The finding which caused such a stir at SMBE 2014 has made it to publication, Altitude adaptation in Tibetans caused by introgression of Denisovan-like DNA. The main result is pretty clear from the figure I’ve placed at the top of this post: it seems that a gene which has been implicated in adaption to high altitudes in Tibetans, EPAS1, exhibits the hallmarks of being derived from the Denisovan lineage. You see that association pretty transparently in the haplotype network illustration. Ed Yong has covered the paper thoroughly at National Geographic. And, he’s done some original reporting, which leads me to putting up that provocative title. This research comes out of Rasmus Nielsen’s lab, and he’s quoted as saying in Ed’s post:
At the time, the world was already populated by other groups of humans, like Neanderthals and Denisovans. As the African immigrants met up with these groups, they had sex. And through these liaisons, their genomes became infused with DNA from people who had long adapted to these new continents. “It’s a new way of thinking of human evolution—a network of exchange of genes between many lineages,” says Nielsen [Nielsen quote in bold].
When I read the above I couldn’t help but think that a “a network of exchange of genes between many lineages” is the dynamic that was supposed to be at the heart of multi-regional hypothesis. Of course there’s some nuance here that needs fleshing out. Classic multi-regionalism implied geographic continuity in morphology, with the core elements of modern humanity emerging through phyletic gradualism across the whole species range. Chris Stringer, who has been a very public face for an “Out of Africa,” or as he’d say “Recent African Orgin” (RAO), model for the emergence of H. sapiens, wrote a review a few months back, Why we are not all multiregionalists now. Obviously by the title you can infer that Stringer is not a convert to multi-regionalism, but, his piece has a rather good survey of the history and nuances of the argument between himself and people such as Milford Wolpoff, and all positions in between. Stringer accepts all the major empirical findings of the genetics enabled first by genomics, and then ancient DNA analysis. To my knowledge so does Wolpoff. In fact Wolpoff feels that the recent findings vindicate his viewpoint, which was relatively marginalized within paleoanthropology of late. In the end to me it seems to boil down to an argument about names, sometimes literally (e.g., should it be Homo sapiens neanderthalensis or Homo neanderthalensis?).
This is not an issue which I’m particularly invested in. For the purposes of understanding the precise details of human evolutionary origins over the past 200,000 years demarcating terminology can be quite useful,* but in terms of general evolutionary relevance, less so. Greg Cochran has put up a post where he notes that he argued on a priori grounds that it was likely that the Tibetan adaptation was introgression years ago. This is true. How did Greg arrive at this hunch? In his most recent post he also allows that Jim Crow had the same intuition. The basic paradigm of adaptive introgression is not particularly novel. Greg told me a long time ago that many results that surprise human evolutionary geneticists would be par for the course in plant genetics, where the idea of local regional adaptations persisting in the face of gene flow are not unknown.
So let’s take a step back here. We now know a few major results which seem robust and have wide support among most scholars focused on human evolutionary origins:
1) ~90 percent or more of human ancestry outside of Africa derives from a population that lived in Africa on the order of 100,000 years ago
2) Eurasian hominins did contribute ancestry to non-African populations, though it seems that genetic incompatibilities had already emerged by the time that Neandertals met the neo-Africans. This is clear in that representation of Neandertal ancestral segments is reduced in genic regions, as well on the X chromosome
3) But, in some regions of the genome the archaic Eurasian hominin representation in modern non-Africans is far greater than expectation, likely due to adaptation
For non-Africans then one might say that in terms of ancestry we are Africans. In terms of most of our functions we are even more African, because the genetic incompatibilities would tend to select against the minority ancestral component, as the background would be dominated by African ancestry. But, in a subset of functions we are far less African due to adaptive value of the archaic Eurasian alleles. Finally, some of these adaptive alleles are strongly regional in their distribution, because their adaptive value is geographically constrained. Current candidates for adaptive introgression such as altitude adaptation and pigmentation seem to clearly fall into the latter class.
Throughout much of this discussion I’ve attempted to put the focus on non-Africans. This was purposeful because it strikes me that we’re coming rather close to the end of the line in the most general sense on the evolutionary genetic origins of non-Africans. There are going to be details in term of ascertaining which genes adaptively introgressed, as well as how many admixture events occurred with how many archaic lineages. But the general shape of the model is in sight. A major neo-African expansion ~50,000 years ago, and a minority uptake of archaic ancestry.
The picture for Africans is somewhat more muddied. The Hammer lab at Arizona has already published on the likelihood of archaic admixture within Africa, and the abstracts from SMBE 2014 imply that they have more coming down the pipeline. But they are the first to admit that their sell is more difficult because of the low probability of recovering high quality ancient DNA from the African remains. In addition I suspect there is also the problem that the admixture scenario within Africa is likely to be far more complex and less clear and crisp than outside of Africa, because of the long history of H. sapiens sapiens and its antecedents within Africa. If I had to bet I’d assume that hominins were more speciose within Africa than outside of it. There may have been repeated gene flow events between “archaic” and “proto-modern” lineages within Africa, to the point where the genetic distance between the two groups was sharply reduced in the first place (methods to infer admixture between divergent lineages might not be able to pick up mixture between populations which had long been engaged in continuous gene flow). Additionally, attempts to reconstruct the demographic history of human lineages often implies that prehistoric Africans did not experience the same magnitude of rapid population growth that non-Africans did, so the stylized model of neo-Africans quickly absorbing a small amount of ancestry from archaic lineages in one go may not be appropriate for Africa. While we may be near the closing of the intellectual frontier in the broadest sense** outside of Africa in terms of recent human origins, within Africa there are still many big picture questions that need to be answered. For a long time African populations, such as the Yoruba, have been featured mostly as an “outgroup” place holder which serves as a baseline and sanity check on the relationships of non-Africans. I assume that in the near future a lot of whole genome analysis is going to come out of Africa, with decent population coverage (yes, Sarah Tishkoff is probably going to be on the author line, so you can get a sense of which populations). The assertion that there is more genetic diversity within Africa than outside is often used to glib effect in my opinion, but in this case I think that this fact may be indicative of future career possibilities for human population and phylo genomicists.
Addendum: Please see this paper as well, Admixture facilitates genetic adaptations to high altitude in Tibet, as it complements the results above. I don’t see it in the references to the more recent paper, and talking to the first author at ASHG 2012 I got the impression she was still skeptical of models of lots of admixture in Tibet between Han-like populations and local indigenes, so that might explain it.
* Also, obviously which side “won” has relevance for the historical assessment of careers.
** I want to reiterate there are many specific questions to be answered. I simply believe that the general framework is starting to come into focus, and we won’t experience many new evolutionary shifts in understanding.