Sunday, July 5, 2015

Revisiting Dobon et al. 2015: The Fulani Samples

I thought it would be interesting to note that the Sahel wondering Fulanis Dobon et al. 2015 sampled were seemingly two or perhaps several distinct subgroups of Fulanis. 

One group for one seems almost homogeneously less West Eurasian than usual for Fulanis in my experience. They essentially seem about as West Eurasian as your average Gumuz whilst the other looks clearly more West Eurasian and pulls left across the X axis toward Southern Levantines, Nagada Copts and Arabians and seem perhaps more representative of other Fulanis like those in Hodgson et al. 2014. [3]

The Sahel

Could this heterogeneity be owed to the fact that when you look at the info shared on the Fulanis; their location is simply listed as "Sahel" (and not a single location/ town/ these were probably wondering pastoral nomads; they in fact are just listed as "nomads" for their "location")? The people in charge of gathering the study's samples probably sampled various distinct Fulanis as a result in my humble opinion but one interesting thing to note is that both the more West Eurasian Fulanis and the less West Eurasian ones demonstrate a clear affinity for Northwest Africans:

Link to the PCA

Virtually all of the Fulani samples began pulling south on the Y axis the same Moroccans, Tunisians & Mozabites do. This is relevant because both they and Moroccan Jews (who also pull south in that PCA above) are known to have non-negligible Northwest African/ Maghrebi admixture. [2] [3]

You see, Fulanis don't get their West Eurasian ancestry from actual West Eurasians like Europeans or West Asians but via proxy from Northwest Africans who themselves are ~70-80% West Eurasian:

This influence from Northwest Africans to Fulanis has been known for a while now and can be demonstrated via ADMIXTURE analyses from studies like Hodgson et al. 2014:

Notice the light green "Maghrebi component" and its substantial presence in Fulanis... It's also interesting to note that David/ the author of the Eurogenes genome blog ran some of these Fulani samples from Dobon et al. through Eurogenes K=8 and they like Northwest Africans/ Maghrebis seemed to have demonstrated Western European Hunter-Gatherer ancestry.

He mentioned that one group essentially lacked such ancestry but another was about 5-6% WHG (these are probably the more West Eurasian Fulanis who pull more to the left on the PCAs I shared and that are ultimately owed to him). [note about the WHG in Northwest Africans]

So it's interesting to note that even those highly African Fulanis are basically still to some small extent Northwest African admixed though they must be barely Northwest African admixed as they cluster more or less as far left as the Gumuz and the Gumuz are very lightly West Eurasian admixed thanks to their seemingly very ancient Omotic admixture:


If they're comparable to that extremely low level of West Eurasian ancestry; I personally wouldn't put their general Northwest African ancestry over ~5% or so but it's clear that it is present given how they demonstrate at least a tiny amount of West Eurasian ancestry and display an affinity for Northwest Africans in a PCA.The other more West Eurasian Fulanis are clearly substantially Northwest African admixed though.

Reference List:


1. Lazaridis et al. 2013 ADMIXTURE run & Harappa world ADMIXTURE run (an example of a run with "Mediterranean" in it)

2. The Gumuz as one compatriot of mine once pointed out likely have a very ancient strain of Omotic admixture, perhaps from a  time when Omotics were far less West Eurasian/ West Asian admixed as they actually have a very substantial amount of "Omotic" ancestry in mostly all runs where Omotic/ Ethiopic shows up but are clearly not very West Eurasian (their est Eurasian ancestry is almost negligible). 

We'll see someday with further study done on various Ethiopian & East African populations & perhaps someday if and when we have ancient genomes from across East Africa.

3. The "Sub-Saharan African" ancestry Northwest Africans display in Eurogenes K=8 is a mix of the East African cluster/ component with some Niger-Congo speaking peoples based African ancestry like what peaks in Yorubas in the higher Ks of many runs like what I mentioned in my blog post about Dobon et al.'s Copts.

4. The "Afar" samples in that last table from Pickrell et al. (reference 4) aren't actually Afars.

No comments:

Post a Comment