search this blog

Loading...

Saturday, March 15, 2014

PCA of ancient European mtDNA


The recent Wilde et al. paper on the ancient DNA of Eastern European steppe nomads included mitochondrial DNA (mtDNA) data for just over 60 of the studied individuals. Below is a Principal Component Analysis (PCA) featuring these samples, marked collectively as KGU, alongside the dataset from last year's Brandt et al. study on the genetic origins of Central Europeans.



Note that KGU falls closest to the Bernburg (BEC) and Unetice (UC) samples from Neolithic and Bronze Age eastern Germany, respectively. This is probably because all of these groups have similar levels of mtDNA haplogroups U5a and H. Moreover, UC is thought to be an Indo-European archaeological culture with origins in Eastern Europe. On the other hand, Brandt et al. hypothesized that BEC might have been of Scandinavian origin.

The Central European metapopulation (CEM) is composed of present-day individuals from Austria, Germany, Poland and the Czech Republic. Its position on the PCA plot suggests to me that modern Central Europeans are largely derived of Kurgan nomads, Bell Beakers from Iberia (BBC), and remnants of Neolithic farmers from the Near East, at least in terms of maternal ancestry.

In other words, I'd say the result correlates well with the findings of Brandt et al., who posited that long-range migrations from eastern and western Europe into the heart of the continent, particularly during the late Neolithic, played an important role in the formation of the modern Central European mtDNA gene pool.

Citations and credits...

Thanks to Eurogenes Project member PL16 for the PCA

Wilde et al., Direct evidence for positive selection of skin, hair, and eye pigmentation in Europeans during the last 5,000 y, PNAS, Published online before print on March 10, 2014, DO:I10.1073/pnas.1316513111

Guido Brandt, Wolfgang Haak et al., Ancient DNA Reveals Key Stages in the Formation of Central European Mitochondrial Genetic Diversity, Science 11 October 2013: Vol. 342 no. 6155 pp. 257-261 DOI: 10.1126/science.1241844

See also...

Extreme positive selection for light skin, hair and eyes on the Pontic-Caspian steppe...or not


7 comments:

sds said...

Interesting David. It appears there was a great deal of migration into Central Europe from all directions. While there is such a diversity of mtdna, I assume the same can't be said of y-dna. Are there any forthcoming studies for this?

Davidski said...

I think the dam will break this year as far as high resolution ancient Y-DNA is concerned, and we'll probably see quite a few results from Central and Eastern Europe. Indeed, that multidisciplinary Swedish project, The Rise, should be coming to an end soon, although the results will probably take a while to publish.

http://polishgenes.blogspot.com.au/2013/04/hundreds-of-prehistoric-scandinavian.html

barakobama said...

That PCA graph is based on HVR1, HVR1, and coding region alleles or just alleles in SNP's right? If so wouldn't back mutations and shared mutations in differnt lineages mess things up? I think just comparing the the results yourself not using a PCA is much better. I don't see any reason for an mtDNA PCA graph.

Davidski said...

The haplogroup assignments are based on "hypervariable region 1 (HVR1) sequences plus 32 mtDNA coding region SNPs and a 9-bp indel".

In any case, you would have a point if the samples on that PCA were all over the place, but they're clearly not. There are some very obvious patterns visible, and generally speaking modern Central Europeans come out looking like a mix of ancient populations from central, western and eastern Europe, which makes sense and is backed up by recent studies.

barakobama said...

"hypervariable region 1 (HVR1) sequences plus 32 mtDNA coding region SNPs and a 9-bp indel".

Whats a 9-bp indel? Is it used the same way as HVR1-2 and coding region?

"In any case, you would have a point if the samples on that PCA were all over the place, but they're clearly not. There are some very obvious patterns visible"

How are the techniques used in this thread better than plainly comparing haplogroup percentages and the circumstances the modern and ancient mtDNA are in. I have not done any serious study on Corded ware, Bell beaker, and Unetice mtDNA but so far they look very similar and happen to also mainly come from the exact same area of Germany and within at most 1,000 years of each other. Bell beaker has obvious Pontaic steppe influence just like Corded ware and Unetice do, and i see little evidence of an Iberian origin. The site with the two R1b samples did not have typical Iberian or Neolithic central European mtDNA, instead i would argue they had mainly Indo European maternal ancestry.

"generally speaking modern Central Europeans come out looking like a mix of ancient populations from central, western and eastern Europe, which makes sense and is backed up by recent studies."

That makes sense but i am not convinced yet. I know almost nothing about modern central European mtDNA so i don't know how they compare to ancient ones. Since Poles and Germans(who are not near the border) have very differnt genetic makeups(proved in Y DNA, autosomal DNA, and even differences in hair and eye color) why would they have mainly the same maternal lineages? Poles and Germans likely have common Loschbour-like and Stuttgart-like ancestors, but western Indo European and eastern Indo European ancestry is probably what makes them the most differnt.

Davidski said...

I don't know much about indels (insertions and deletions), but yes, it just seems like another way to characterize mtDNA sequences.

But in any case, I think it's pretty obvious that the Bell Beakers expanded from the Atlantic fringe, and then mixed in Central Europe with other groups coming from the east.

Also, what makes Poles and Germans different is that Poles have more ancestry from the Balto-Slavic and Slavic expansions. Many Germans have that influence too, but not to the same extent.

Indeed, much of the genetic differentiation you see between Europeans, especially across the north of the continent, is mostly due to these recent demographic events, particularly the founder effects linked to the expansions of linguistic and ethnic groups that we know today. A lot of people don't understand this, including many population geneticists.

barakobama said...

"Also, what makes Poles and Germans different is that Poles have more ancestry from the Balto-Slavic and Slavic expansions. Many Germans have that influence too, but not to the same extent.

Indeed, much of the genetic differentiation you see between Europeans, especially across the north of the continent, is mostly due to these recent demographic events, particularly the founder effects linked to the expansions of linguistic and ethnic groups that we know today. A lot of people don't understand this, including many population geneticists."

I don't have much to say about that yet but it's a very interesting theory. The Rise project will defintley be useful to figuring out the origins of modern northern Europeans(probably not Finno Urgics though). That statement reminds me of how it seems light hair hasn't been able to find an ancient source(Indo iranians-) with ancient DNA.

Maybe high amounts of common ancestry in northern Europeans can explain why probably recent(based on ancient DNA) light pigmentation is so widespread and appears to be very ancient. I have a mixed opinion right now of where it comes from, and if its very recent like Neolithic-copper age there had to be some very crazy evolution and population replacement(maybe Kurgen Siberians came from the same light source). There are some inconsistencies though with the very recent Neolithic-copper age theory.