Thursday, June 30, 2016

PCA & ADMIXTURE results for Natufians

That's a 3D interactive PCA (Principal Component Analysis) based on autosomal SNPs made by David Wesolowski who authors the Eurogenes genome blog and ancestry project. What's particularly interesting to me about it are the PCA positions of the Natufians and the Neolithic Levantines... With the former group pulling southwards toward African populations such as North, East & West-Central Africans. 

Eurogenes ANE K7

ENF: 77%
East African: 8%

Hunter_Gatherer vs. Farmer

Middle Eastern Herder: 64%
Mediterranean Farmer: 30%
East African Pastoralist: 7%

Eurogenes K12b

Southwest Asian: 54%
Mediterranean: 38%
East African: 8%

That pull along with the above ADMIXTURE results (via Gedmatch) of one Natufian seem to contradict what Lazaridis et al. was saying about the Natufians lacking African admixture but I would caution against using modern PCA positions (like those of Bedouins) and, of course, modern ADMIXTURE runs (with modern clusters based on modern genetic diversity) to gauge how "African" or "Eurasian/Out-of-Africa" an ~11,000-14,000 year old population was.

I.e. These Natufians are, of course, not "Southwest Asian" + "Mediterranean" but, instead, they're just showing the greatest affinity for these modern clusters. As in, populations probably quite like them to some degree; contributed to the formation of clusters like Southwest Asian & Mediterranean.  But, it's still strange that they'd show such an affinity for an African cluster like the East African one.

"However, no affinity of Natufians to sub-Saharan Africans is evident in our genome-wide analysis, as present-day sub-Saharan Africans do not share more alleles with Natufians than with other ancient Eurasians (Extended Data Table 1). (We could not test for a link to present-day North Africans, who owe most of their ancestry to back-migration from Eurasia)."

If anything, it makes what Lazaridis et al. noticed with their analyses, as noted in the above quote, all the more interesting. It could, in my opinion, just be a quirk of their age. Mota, according to the academics who sampled him, didn't have any "Eurasian" admixture based on formal statistics (from what I recall) yet his PCA position (pulling toward Out-of-Africa populations more than the Southern Sudanese or Western Africans) on a global PCA implied otherwise:

Mota's ADMIXTURE results also implied some vague and broadly Out-of-Africa/Eurasian admixture:

Eurogenes ANE K7:

ANE: 3%
ASE: 2%    
East Eurasian: 2%
West African: 20.30
East African: 65.23
ENF: 7.65

Hunter_Gatherer vs. Farmer

Baltic Hunter Gatherer: 1%
South American Hunter Gatherer: 1%
South Asian Hunter Gatherer: 2%
East African Pastoralist: 56%
Oceanian Hunter Gatherer: 1%
Pygmy Hunter Gatherer: 25%
Bantu Farmer: 14%

Eurogenes K12b

East African: 54%
West African: 40%
South Asian: 2%
Siberian: 1%
Western European: 1%
East Asian: 1%

And Mota is only a ~4,500 year old Southwestern Ethiopian sample. So, I'd remain rather skeptical about what global PCA positions, where these pre-historics are thrown in alongside modern populations, have to say. And ADMIXTURE runs that were usually (not always) based on modern populations can also produce somewhat dubious results.

The fact that they lack African admixture (discounting whatever Basal Eurasian's cause will one day turn out to be) may still be quite the case if they truly do not share more alleles with African populations than other pre-historic Out-of-Africa samples do (such being Lazaridis et al. 2016's finding).

On another note, David's placement of them in his mostly Pan-West Eurasia PCA is rather intriguing as well. In this case, they do not cluster with any modern West Eurasian populations and; like NW Neolithic Anatolians, Early European Farmers and Sardinians; they break off from other West Eurasians as they seemingly lack the eastern (it's "northern" in this PCA, I suppose) pulling affinities (I.e. ANE-related admixture) which somewhat pull all the other populations away from them. 

The Early European Farmers and Neolithic Anatolians pull much more west toward WHG/Villabruna-type peoples while the Natufians cluster just south of Negevite Bedouins which implies that they have the least "WHG" related affinities when compared to the Neolithic Levantine, Anatolian and European samples. Which is, roughly, what Lazaridis et al.'s ADMIXTURE run implied:

It really does, as I pointed out earlier, look like these Natufians (and their Neolithic Levantine counterparts) might, for the most part, be the modern Southwest Asian cluster or David's old ENF/ Near Eastern cluster (or Lazaridis et al. 2013-2014's Near Eastern component) in the flesh. The Neolithic Levantines, in the Pan-West Eurasia PCA, just look like somewhat WHG-shifted versions of the Natufians. Perhaps this shift is why they are supposedly about ~15% less Basal Eurasian than their Natufian predecessors:

At any rate, I'll leave it at that for now... I'll be interested in seeing what analyses from third parties like David such as d-stats turn up on these Natufians (haven't really sifted through those yet).

Reference List:


1. Gedmatch kit number for the Natufian: M041601 AND the Gedmatch kit number for Mota: M261275

2. I'm open to these Natufians having African ancestry. Just pointing out that PCA positions and the current ADMIXTURE runs we have might not be the way to go. Nevertheless, it is interesting that these Natufians and not the later (~7,500 year old) Neolithic European samples like Stuttgart (F999916) show such African affinities in ADMIXTURE runs. Might just be because of the very heightened WHG-shift in a Neolithic European Farmer like Stuttgart and perhaps the sample's younger age by ~4,000-7,000 years.

3. Mota's Eurasian admixture in those calculators also, unlike the Natufians' African admixture, looks more broad (I.e. ANE + East Eurasian + ENF etc.) and looks more like an affinity. The Natufians, on the other hand, seem to mostly lean toward the East African cluster which looks less like a broadly African affinity and more like actual admixture, quite frankly.

4. Most of the Natufians do clearly have a bit of a western-shift though. They don't cluster exactly where David's "ENF" cluster theoretically would (only one of them does).

No comments:

Post a Comment