May 12, 2011

The origin of Sorbs

An important caveat from the paper:
One caution regarding our results is that the geographical origins of our reference populations are crudely characterized only by country and thus may not be random samples. If many of the Germans in the POPRES data are western German samples, this may inflate the apparent differences we observe between Germans and Sorbs. The LPZ Germans contained two individuals from Eastern Germany who do appear closer to the Sorbs, suggesting that population structure within countries is a valid concern. Certainly, a tighter and denser sampling of German, Polish and Czech individuals from regions surrounding the Sorbian territories would be ideal for confirming or refuting the results found in this study.
Another important point:
The MAF spectra (Supplementary Figure 6), although highly distorted because of SNP ascertainment, also show the Sardinians and Basques to have a noticeable excess of monomorphic SNPs. This excess suggests that some SNPs that are polymorphic in Europe may have been driven to extinction/fixation at a higher rate or never existed at all in these populations, consistent with genetic isolation.
It is clear that at least part of this phenomenon is explained by the fact that some polymorphic SNPs in non-Basque and non-Sardinian Europeans are due to Eurasian influences that these two populations lack. In small isolated populations, alleles may be driven to fixation, while larger populations may continue to exhibit polymorphism at the same loci. However, that phenomenon would not preferentially shift either Sardnians or French Basques away from East Asians. Hence, rather than think that polymorphism was lost in them, we can think that polymorphism was added to non-isolated Europeans due to phenomena along the east-west Eurasian axis.

European Journal of Human Genetics
advance online publication 11 May 2011; doi: 10.1038/ejhg.2011.65

Genetic variation in the Sorbs of eastern Germany in the context of broader European genetic diversity

Krishna R Veeramah et al.


Population isolates have long been of interest to genetic epidemiologists because of their potential to increase power to detect disease-causing genetic variants. The Sorbs of Germany are considered as cultural and linguistic isolates and have recently been the focus of disease association mapping efforts. They are thought to have settled in their present location in eastern Germany after a westward migration from a largely Slavic-speaking territory during the Middle Ages. To examine Sorbian genetic diversity within the context of other European populations, we analyzed genotype data for over 30 000 autosomal single-nucleotide polymorphisms from over 200 Sorbs individuals. We compare the Sorbs with other European individuals, including samples from population isolates. Despite their geographical proximity to German speakers, the Sorbs showed greatest genetic similarity to Polish and Czech individuals, consistent with the linguistic proximity of Sorbian to other West Slavic languages. The Sorbs also showed evidence of subtle levels of genetic isolation in comparison with samples from non-isolated European populations. The level of genetic isolation was less than that observed for the Sardinians and French Basque, who were clear outliers on multiple measures of isolation. The finding of the Sorbs as only a minor genetic isolate demonstrates the need to genetically characterize putative population isolates, as they possess a wide range of levels of isolation because of their different demographic histories.



eurologist said...

Pretty much as expected.

However, I always find the use of faraway countries in PCA analyses strange, useless, and detrimental to the best possible research outcome. I would have restricted samples used to Scandinavia, Netherlands, Belgium, Germany, Austria, Hungary, and a wide set of Slavic countries, perhaps including Balto-Slavs.

pconroy said...


I would add Poland and Ukraine to your list, as that's where the Sorbs originated...

arvydas said...

eurologist: "...and a wide set of Slavic countries, perhaps including Balto-Slavs.\"

You sound as if there were Slavic countries AND Balto-Slavs. That leaves me wondering who might be those mysterious "Balto-Slavs"...


Nirjhar999 said...

Good job.

eurologist said...

I commented on this before, but that got wiped out by the latest Google crash.

My main thrust was that the results are as expected. I detest the use of populations from far-away regions to shed light on a clearly local phenomenon. Yes, I understand some of the effort was to quantify the degree of isolation (which is certainly not expected to be that huge over just ~1,000 years), but come on - for the PCA analysis, it is much more informative to use data from the nearest 500 - 1000 km, or so. We don't have to include all of Europe (or neighboring Asia)...

More diversified and cleanly located data within Germany and adjacent Slavic countries is needed to identify when these populations mixed, and when they were separated. From y-DNA, there are already glimpses that at the onset of agriculture (or even before), certain R1a populations were wide-spread - but those of Slavic Expansion are entirely different (and may not make up more than 10% to 20% in much of the northwestern portion of eastern Europe).