This model has recently been shaken by the discovery that some modern humans are closer to extinct hominins: the Neandertals and Denisovans. Within SROOA this is quite unexpected, as these groups were thought to be irrelevant side branches of the human family tree: differential affiliation of some modern humans to them means that they are not.
This has invariably led to an acceptance that SROOA is basically wrong in its purist form, and an alternative form including some assimilation of archaic hominins is moving to become the new default position.
Now, a new paper that works within the SROOA tradition itself finds another surprising piece of evidence that contradicts SROOA. But, first, let's examine what SROOA entails:
- A human group (Proto-Eurasians) moved Out of Africa, starting to become differentiated from Africans
- At a later time, different Proto-Eurasian groups became effectively separated by geography and started diverging from each other, eventually evolving to become Europeans, East Asians, Australian Aborigines, etc.
- Under this model, different Eurasian groups are expected to have exactly the same position vis a vis Africans, once drift is accounted for. Each population's divergence from Africans can be analyzed into two components, e.g., an African-Proto-Eurasian one, and a Proto-Eurasian-Chinese one. The genetic drift (away from Africans) in the first part is the same for all non-African populations (e.g., Chinese or Tuscans), while in the latter it is dependent on population history.
I had noticed this phenomenon in my experiments for quite some time, and highlighted the possibility that haplogroup E1b1b, which is shared by West Eurasians and Africans may have something to do with it.
If autosomally African-like populations moved out of Africa carrying these Y-chromosomes, then we would expect only E1b1b-bearing populations to be shifted towards Africans relative to the Chinese. The authors rightly reject this idea because they discover that all European populations, including a wide variety of North European ones, where haplogroup E is non-existent, are closer to Africans than the Chinese are.
So, we are left with a bit of a problem:
- There is clearly something "common" between West Eurasians (even North Europeans) and Sub-Saharan Africans that is not shared by East Asians, a factor X which brings them closer to each other than would be expected by the SROOA
- The best candidate of this "common" element (Y-haplogroup E) does not have the expected distribution.
Solution #1: West Eurasian Back-Migration into Africa
The first solution is based on the idea that there has been a major episode of back-migration into Africa that is not captured by the standard model. Clearly, this cannot have been a recent event, as Sub-Saharan Africans largely lack (except in the North and East) certifiably West Eurasian derived markers. But, the event need not have been particularly recent: as long as it occurred after West Eurasians began to diverge from East Asians it would have established the genetic closeness observed by the authors.
I believe that the best signal of such a potential back-migration involves haplogroup E. This is the dominant patrilineage of black Africans by far, and almost certainly had an African (and probably an east African) origin. However, its sister haplogroup D occurs as a relic in Asia, among people such as Tibetans, Ainu, or Andaman Islanders. Where did the ancestral clade DE develop? If I was a betting man, I would say that somewhere between the Indian Ocean (where the Andamanese live), and East Africa.
An early movement of DE-bearing men from Arabia into East Africa would serve to bring a West Eurasian autosomal component into Africa. That component would then evolve into E in East Africa itself, and go on to (almost) completely replace pre-existing African Y-chromosomes, leaving haplogroups A and B at high frequencies in a few relic African hunter-gatherer populations.
From the Eurasian perspective, the problem would evaporate: West Eurasians' autosomal shift to Africans is not correlated with haplogroup E frequencies, because the latter was not initially associated with a Sub-Saharan-like autosomal gene pool.
Solution #2: Archaic admixture in East Asians
A second potential solution would interpret African-West Eurasian closeness not as evidence of a common population element in West Eurasians and Africans, but as a consequence of a population element in East Asians that both West Eurasians and Africans lack.
The obvious candidate for such an element involves archaic admixture in East Asians. This is no longer an exotic possibility, given the evidence for such admixture that the sequencing of the Neandertal and Denisovan genome has produced.
Under this scenario, Europeans and East Asians are genetically close because of their common Proto-Eurasian ancestry, and they both shared an initial same distance to Africans, but East Asians diverged from both West Eurasian and African populations by admixing with archaic humans they encountered in the East.
A test of solution #1
Suppose that solution #1 is correct. Consider also that:
- West Eurasians are shifted relative to East Eurasians by x% to San Bushmen (who have some of the highest frequencies of non-E chromosomes), and are expected to be least affected by any sort of West Eurasian->Africa back-migration. C
- West Eurasians are shifted relative to East Eurasians by y% to Yoruba (who have extremely high E-haplogroup frequencies), and are expected to be more affected by West-Eurasian->Africa back-migration than the San are.
It is not clear to me whether East Asian archaics/West Eurasian back-migration into Africa, or a combination of these factors may account for the observed phenomenon. I should point out a second Out-of-Africa expansion into West Eurasia (the authors' preferred model) is also not out of the question, but this cannot be easily tied to a particular event in prehistory or uniparental marker.
The study of human origins has just gotten even more interesting...
Genome Research doi:10.1101/gr.119636.110
Human population dispersal “Out of Africa” estimated from linkage disequilibrium and allele frequencies of SNPs
Brian P. McEvoy et al.
Genetic and fossil evidence supports a single, recent (less than 200,000 yr) origin of modern Homo sapiens in Africa, followed by later population divergence and dispersal across the globe (the “Out of Africa” model). However, there is less agreement on the exact nature of this migration event and dispersal of populations relative to one another. We use the empirically observed genetic correlation structure (or linkage disequilibrium) between 242,000 genome-wide single nucleotide polymorphisms (SNPs) in 17 global populations to reconstruct two key parameters of human evolution: effective population size (Ne) and population divergence times (T). A linkage disequilibrium (LD)–based approach allows changes in human population size to be traced over time and reveals a substantial reduction in Ne accompanying the “Out of Africa” exodus as well as the dramatic re-expansion of non-Africans as they spread across the globe. Secondly, two parallel estimates of population divergence times provide clear evidence of population dispersal patterns “Out of Africa” and subsequent dispersal of proto-European and proto-East Asian populations. Estimates of divergence times between European–African and East Asian–African populations are inconsistent with its simplest manifestation: a single dispersal from the continent followed by a split into Western and Eastern Eurasian branches. Rather, population divergence times are consistent with substantial ancient gene flow to the proto-European population after its divergence with proto-East Asians, suggesting distinct, early dispersals of modern H. sapiens from Africa. We use simulated genetic polymorphism data to demonstrate the validity of our conclusions against alternative population demographic scenarios.