Some interesting fresh aDNA from Central Europe

I recently came across a new publication: Krause-Kyora e.a. “Neolithic and medieval virus genomes reveal complex evolution of Hepatitis B”.

The main message sounds familiar: Hepatitis B viruses present in Central Europe during the EN/MN went extinct, to be replaced sometimes between the LN and the High Medieval with the strain haplogroup that is dominating Europe to date. When exactly, and how that happened requires further analysis. Apparently, more studies of this kind are underway and may one day provide a more detailed picture.

Krause-Kyora e.a. 2018, Fig. 1 : Origin of samples

Somewhat hidden, and poorly discussed, the paper also provides 3 new samples of human aDNA that deserve further exploration. Data has been uploaded to the European Nucleotide Archive, accession No. PRJEB24921.

The earliest of these samples, LBK, Karsdorf (Saxony-Anhalt), 5056–4959 cal BC,  clusters closely with other LBK samples, so there is not much surprise to be expected from in-depth analysis.

Things are different with the next one, TRB-Tiefstichkeramik, Hildesheim-Sorsum, 3335–3107 cal BC:  It closes an important geographical, chronological and cultural gap, and holds several surprises – enough to devote a separate posting to it.

The Petersberg (Source: Wikimedia)

The last one is from the Petersberg Abbey (SE Bavaria), 1020–1116 cal AD. The Abbey, located on the German-Austrian border, overlooks the Inn valley where it leaves the Alps. Given that strategic position on a major North-South corridor (today’s E45 Norway-Sicily), the sample’s character is also unexpected. It clusters amidst Unetice, and certain BB samples from Augsburg, proposing a surprising extent of population continuity between the BA and the High Medieval. In terms of modern populations, the sample comes across Western-Balkans like, with the closest modern populations being Bulgarians, Albanians and Tuscans, plus a single Basque (see the PCA below).

Krause-Kyora e.a. (2018), fig. suppl. 9 (modified): Principal Component Analysis (PCA) of the human Petersberg sample projected on 27 modern day West Eurasian populations

Contrast this with the early Medieval (Migration Period) Bavarians reported in the recent K. Veraamah “Elongated Skulls” paper, which were found only some 75-150 km away from the Petersberg. Those that not posessed elongated skulls came out as tendencially NW European (“Frankish”).

Veeramah e.a. (2018), Fig. 4: Geographic distribution of population assignment analysis (PAA) results (..) for (A) all Bavarian males, (B) all Bavarian females with normal skulls, (C) all Bavarian females with elongated skulls, and (D) KER_1 and VIM_2

What may explain this contrast?

  • Genetic pull towards the E. Balkans, effected by the “females with elongated skulls” analysed by Veeramah ? I wonder whether the latter were numerous enough to bring about such a strong “pull”. Moreover, would admixture between W. Germanic (Frankish) and E. Danubian (Gothic/ Alan) populations really have brought about a population that appears virtually indistinguishable from Unetice?
  • Slavic immigrants? The Bavaria Slavonica is well documented, but mostly confined to E. Franconia, plus certain monasterial areas further West, where Slavs were called in to colonise vacant farmland (“free”, weapon-bearing Germanics were typically unwilling to subdue to monasterial control and taxation, apparently, this was less of an issue with immigrating Slavs). However, for the Inn valley, where the Petersberg is located, no Slavic settlement is documented or indicated by toponymy. Moreover, (descendents of) Bavarian Slavs should plot close to modern Czechs, which is not the case with the Petersberg sample in question.
  • Base map by EuRegio Inntal, modified
    Long(er)-term population continuity? IMO, the most likely scenario. The region in question holds quite a number of pre-Germanic toponyms. Madron, the alternative name of the Petersberg, e.g., is believed to derive from Celtic mat “big, high” & rin “mountain” (c.f. the Rhön mountains in N. Bavaria / SE Hesse). Erl/ Tirol, less than 10 km to the SE, is acc. to P. Anreiter, Austria’s most reknowned toponymic researcher, originating from a Roman (Villa) Aurelia. And a bit further south follows Ebbs/ Tirol, first recorded in 788 CE as ad EPISAS. Episas is being connected to the Celtic horse goddess Epona and interpreted as “horse creek”, and, in fact, through Ebbs flows the Rossbach (“horse creek”). Further up the Inn, between Wörgl and Innsbruck, there are even various pre-Celtic (para-Venetic) toponyms preserved.

Hence, the Petersberg sample appears to demonstrate a substantial degree of population continuity in the middle and lower Inn valley, in spite of being a major transit corridor, and Celtic (La Tène), Roman and Germanic immigrations/ conquests during the IA and the Early Medieval. The Petersberg was already inhabited in the Bronze Age (Urnfield and LBA) [it even contains a Late Paleolithic burial]. I don’t know whether recent excavation there provided BA bone material that may yield aDNA. If so, I would be curious about the results – did Unetice-like populations already settle there during the Urnfield period ?

7 thoughts on “Some interesting fresh aDNA from Central Europe

  1. Thanks Frank.

    It will be interesting to see further research about the possible impact of disease in the European LN/EBA transition with the already well known genetic shift that happened at that time. There is interest in this field, not just for historical reasons, but probably also medical ones.

    Regarding the Pertersberg sample, I’d not that the PCA with modern populations projected over the ancient ones looks a bit deceiving. It can’t be that modern North Europeans cluster with the Mesolithic HGs, so I trust more the ancient to ancient correlation, where this sample looks like Unetice and BB, and therefor it should cluster with modern Central/Northern Europeans too. We’ll see when someone gets the genomic data.

    Looking forward to the TRB-Sorsum sample post, as that sample is the most interesting one as you noted above.

  2. Frank I look forward to your thoughts on TRB, especially when you say it “closes a missing link”
    We have a couple of TRB and MN Chamber grave samples from central-north Germany, but still a lot missing – especially the big gap between these post-4000 BC samples and LBK, which ended 5000 BC. So we are looking at a 1,500 year gap in genome-wide aDNA (there are some mtDNA data).
    So I am interested to see what happens after the demise of the LBK, with follow-up groups like Rossen, Gatersleben, etc

    Of course, also from west Germany, and data from France, esp Michelsberg.

    About Bavaria, my feeling is that there was considerable migration post 450 AD (with the final collapse of Roman control), from Bohemia – the Friedenham-Prestovice group (which then vacate southern Bohemia), as well as migrants and “refugees” from Alemannia & Thuringia fleeing Frankish incursions c. 500 AD.

  3. Great to see Alberto, Rob, Kristiina and FrankN have gathered here.

    When will FrankN’s separate posting about the tantalisingly described “surprises of the TRB-Tiefstichkeramik, Hildesheim-Sorsum, 3335–3107 cal BC” appear?

    Are we allowed to make requests for further guest post topics?

    Could Kristiina be requested to write a guest post or series on where she thinks the Finno-Ugric homeland is, when its first speakers were in Europe, and her reasoning for both? What are her current views on whether and how Hg N is connected to the language family’s earliest entry and distribution in Europe? Do only some groups of Finno-Ugric speakers have notable ancestry from Yamna or is that ancestry generally shared by speakers of the overall language family, including among Finno-Volgaic and Finno-Permic speakers? When may this have happened and why did it not result in a language change to Indo-European? Or was it a language change from Indo-European to Finno-Ugric languages? And if so, what events or technologies may have caused this?

    I am also interested to know if Kristiina favours any homeland for Turkic and why. What is the nature of the relationship between Mongolic and Turkic languages (or their ancestral forms), since there is some historic association between the population of speakers of both, despite the language families being distinct? Is this association due to a meeting of the two groups of speakers at some point in time, or may there be a genetic connection deeper in time? (I’m not proposing the Altaic language family, but rather asking for Kristiina’s views based on her background in linguistics.) How may Turkic have formed?

    Do Kristiina, FrankN Alberto, and Rob have any views, based on aDNA or otherwise, on the homeland for Basque and its ancestors? And if any considers Basque migrated to its current position, when may this have occurred?

    Would all four write in more detail about where they each think the PIE homeland might be and their detailed reasoning (especially if any homeland outside the steppe is favoured), based on the current aDNA available as well as their knowledge of archaeology, linguistics and all else relevant?

    Will there be more in-depth discussions, particularly by the above 4 people, about the two Southeast Asia aDNA papers, highlighting anything each person found of particular interest there? I was somewhat disappointed at the lack of deeper discussion of these 2 papers at anthrogenica and eurogenes.

    What does each person think about Y-DNA Hg T and how it may have attained its current distribution? Are there any indications how and when L and T may have come by their different distributions?

    Do any among Alberto, Rob, Kristiina and FrankN have predictions or expectations about the genetic composition of Hittites and of Mycenaean and other ancient Greeks, if they are not convinced that steppe ancestry may have been a major contributing factor in either the ethnogenesis of the speakers or the languages themselves? In each case, what is the reasoning leading to such expectations?

    (I also had some specific questions, a few of which are from a while back, but have to organise them.)

  4. @ak2014b

    Hi, and welcome. Those are a whole lot of questions and you still have more! Let’s see if I can answer at least a few of them.

    I have indeed invited the three persons you mention to write some guest post whenever they feel that the comment areas or a forum are not enough for what they’d like to express. And so they can also keep a place with some of their writings. But otherwise they’re not specifically affiliated to this blog in any other way. My posts might have ideas they don’t share at all, so it’s not a collaborative effort in that sense.

    I do hope that Kristiina will write up someday her thought about Uralic and its relationship to IE or Turkic. But I can’t answer if or when she will do it (or she might choose a different place to do it). In any case, whenever there are any news regarding those topics I’m sure she’ll be around to comment about them.

    FrankN’s second post is coming soon. If not this week, then I hope early next week. We’ll see.

    For the rest of the topics I’m sure we’ll discuss them here sooner rather than later. I had to catch up with a few things when I moved here and my time is limited, so I still didn’t find the time for the SE Asian paper (which also is not about an area I know quite enough about to have much to say, but something will come).

    Mycenaeans and Hittites will surely appear here soon, as they’re becoming quite central to the Indo-European question that aDNA is trying to solve. Also I expect that soon we’ll see the Rakhigarhi paper, so stay tuned for that one too.

    It’s summer, so things are a bit slow. But I hope that soon they will start to take off again. Hope to see you around!

  5. @ak2014b

    In preparing the Sorsum post, I reviewed some very recent German & Scandinavian publications that, in short, suggest that the genesis of the TRB phenomenon may have been much more complicated than assumed so far. This has delayed me a bit, but the Sorsum post should latest come early next week.

    In the process of this review, it also became apparent that a fresh look at some of the contemporary “well-known” aDNA, e.g. Baalberge and Esperstedt_MN, might be required. Alberto, Rob and myself are currently internally discussing respective details, and one or several of us may sooner or later post about it,

    As concerns the IE homeland, my ideas about a possible location in NW Iran have, I think, become already clear in my postings on Eurogenes. I intend to detail that further by
    (a) highlighting innovations that during the late 5th mBC spread out of NW Iran via Maykop into the Steppe and beyond (no – not horses and wheels, rather wool sheep and non-light sensitive barley, for both of which DNA analyses are available);
    (b) a closer look at the role of Kura-Araxes in linking Central Asia and the Mediterranean, that may also relate to the Hittites/ Mycenean question;
    (c) reviewing linguistic indications for IE-Sumerian language contact (pre-Mittani), which IMO cannot be explained via the Steppe theory [I may ask Kristiina for help on that one.]

    Otherwise, there is S. Nikolaev’s recent proposal of Proto Wakashan-Nivkh-Algic (PAW), another trans-Pacific language family after Aleutic and Yenissean – Na Dene. With the Algic homeland assumed to be around Kennewick, and the ANE/ Kennewick Man admixture recently discovered in Steppe Maykop, this language family is of more than more academic interest. I spent some time in comparing Swadesh-100 roots for PIE, PUralic and PAW, with the surprising result that PIE seems to be closer to PAW than to PUralic (well, not so surprising – specific closeness between IE and Chukotko-Kamchatkan [incl. Nivkh] had already been demonstrated previously). You’ ll get a detailed post about this sooner or later. I am not sure yet how to explain the results, but, if real, they should clearly not only be relevant for the PIE homeland debate, but also as concerns the Uralic and Turkic homelands.

    Finally, SEA. I have worked and lived more than two years in the region, and read the papers with a lot of interest. I think they are well-made, but of course just a first step towards disentangling the region’s prehistory. I hope there is more to come.
    AFAIK, having learnt that there is interest in SEA, I may one day write something here about the “India-sation” of the region, especially in connection to Bronze metallurgy and the Burmese and Malaysian tin deposits, but also the spread of Hinduism and Buddhism, and the linguistic influence of Dravidian and Sanskrit. Ideally, however, we should first have aDNA from India to help sorting out to which extent cultural influence also left genetic traces (btw – also in the opposite direction: Bananas, e.g., were an IVC import from Insular SEA.)

  6. Thank you both very much. I’m looking forward to reading everything you will be writing, whether it is on the topics I asked about or of any other areas of interest to you.

    FrankN, will you eventually be contributing a guest post going into further details about PIE and PAW? That sounds very new to me. And I’m also looking forward to what you may have to say about Southeast Asia, whether it be “India-sation” or the periods before.

    I had a couple of questions related to horse domestication and chariots for both of you and for Rob. Alberto’s new post on “Horses and Wheeled Vehicles” comes at the perfect time, so I’ll try to formulate my questions there.

  7. I have located my old question for Rob.


    Last year I’d posed a question to you on eurogenes ( about a comment you’d left at Your comment and Davidski’s own response got removed. And all my subsequent comments with questions for you referring to your earlier statements were repeatedly removed as well (I’m still not sure why). So you may not have got to see my question then. Here’s my chance to ask it again, as I’d saved it.

    “I wanted to ask Rob whether he made the following statement based on his early knowledge of upcoming aDNA papers.

    ‘Rob said…

    But it seems I2a2 is the least of your worries. How you going to explain the R1a in Neolithic India ?
    May 24, 2017 at 4:29 AM’

    But David replied with equal certainty as follows, so I concluded maybe they each heard of results from 2 different aDNA papers from the region:

    ‘Davidski said…


    And there’s no R1a in Neolithic India.
    May 24, 2017 at 4:32 AM’

    I recall that Rob said over at anthrogenica that he has contact with archaeologists working on various papers, and have assumed that that’s where he may have learned of this.

    The beginning of the Chalcolithic in the Indus Valley is supposed to be contemporaneous with the same era in Iran, which I remember being mentioned as 6th or 5th millennium BCE. If true, then the Neolithic in the region would be before that. Rob, is the R1a you’ve heard of from there indeed from the local Neolithic?”

    I now have further questions to add to the above: since your statement came across as more certain than guessing, are you at liberty to reveal which study or paper this sample is part of? Is it a paper that has already been published or one that has yet to appear? And is it of the same study or studies you alluded to over at anthrogenica around late 2016, when you referred a few times to South (Central) Asian aDNA results being mostly ready by then and that “a handful of people know the results already” ( and


Comments are closed.