June 27, 2012

Population structure in Qatar

The recent publication of Omberg et al. (2012) has reminded me of the data of Henn et al. (2012) on Qatar which I don't believe I've used yet. I used the K12b calculator on ~20,000 SNPs that are common between it and the Affymetrix chip used.

Below is the population portrait of the Qatari population:
Obviously this isn't a homogeneous population. In order to figure out which ancestral groups are present there, I ran MCLUST over the admixture proportions, which resulted in individuals assigned to five different clusters. Here are the average admixture proportions of these five clusters:

On the basis of the above, I conclude that there are several different groups represented in the Qatari population. I have absolutely no knowledge about the Qatari population, so it would be interesting to see if readers find correspondences between these and known social divisions in Qatar.

For example, I could wager that #5 which is a "Southwest Asian"+"Caucasus" mix represents a pure Arabian group with little outside influences. #1 and #2 are also Arab-like but with various degrees of admixture. #3 appears to include substantial African descendants and #4 a clear Iranian signal due to the high "Gedrosia" component. Of interest is that the "African" group #3 also scores high in the "South Asian" component.


Maju said...

"Of interest is that the "African" group #3 also scores high in the "South Asian" component".

Reality check needed.

Looking at the individual bars: those who have lots of South Asian have no or very little African and vice versa. Your statistical finding is wrong therefore.

AdygheChabadi said...

Actually, Maju, it may be accurate if these people were Siddis from India...

Siddis are East Africans by descent brought to Persia and the Western Sub-Continent by various slave traders...Portuguese, Persian, and Arab.

There was a major genetic study on them recently that Dienekes posted about.

Katharós said...

The level of Admix is not necessarily surprising in that region if you view the seafaring tradition of the area ever since the Bronze Age, like Majan”Oman” and its links to the Indus Valley Civilization. Or Herodotus claims on Phoenician origins.
Or take the Arab-Persian tale of Sinbad the Sailor, the name Sinbad also refers to the land of the Sindhi people”Pakistan”

On the other hand the term “vomit of the sea” is sometimes applied to people living along the coasts based on there blurry "Arab" origin.

eurologist said...

Clusters #3 and #4 are most interesting, to me. They appear to be the least indigenous (by far the least SW Asian). #3 makes this up with almost 1/2 African, and 1/6 South Asian. So, I would think slave trade, African workers, and/ or sailors.

#4 Is different in that here, Caucasus and Gedrosia (i.e., Anatolia to Pakistan) make the main contribution, with significant N European levels. Looks like this diametrically opposed to the above, but an ancient northern component made it all the way to Qatar, without much of a problem. Polygamy, slavery, and harems may have played a role.