No announcement yet.

Reference Populations and Methodologies for detecting ancient Celtic ancestor

  • Filter
  • Time
  • Show
Clear All
new posts

  • Reference Populations and Methodologies for detecting ancient Celtic ancestor

    I realize it is not an exact science and depends on the reference population, but my results from myHeritage in comparison to my results from ftDNA do not look right. The former seems off, the latter seems more accurate based on my pretty extensive family tree.

    The reason I'm confused by myHeritage is that both of my parents are from the same part of NW Spain (Southern Galicia), and the families grew up within less than 100 km / 60 mi of each other.

    In the post neolithic era, the place was settled by Celts, who resided there for centuries. While the language is gone, the people (including the northern Portuguese) remain fairer skinned than the rest of Iberia, there is some remaining mythology, and the folk music, dance, and instruments are still very Celtic, e.g. we play have bagpipe pipers at weddings and some people know how to folk dance similar to river dance.

    My cousin has compiled a pretty extensive family tree, and with the exception of a 3 x Great Grandparent who may have not been local, we can go back 5 - 9 generations, it's all Galician Spaniards, and the next generation or two that we don't have because they predate records can be assumed to be Spanish on both sides as well by the Paternal Maternal styled surnames.

    This is a poor area, so there was not a lot of gene transfer in since the Romans to the Visigoths in the 3rd to 6th centuries. The Moors spent little time there.

    That's a long way of saying I would expect to get 'Celtic'-related hits in my ethnicity estimates, along with possible Germanic, and probably north African.

    Here are my results:


    72% Iberian
    17% British Isles
    6% Eastern European
    5% North African

    That's pretty much what I would expect. I match reference populations in GB, and GB was also settled by the Celts. GB, like Spain, has also had Germanic influence. Spain had the Suevi and then the similar Visigoths, who were western Germanic. I'm not that great on British history, but I know the Aengles, the Anglo-Saxons, and other groups were Germanic, and obviously the Irish and Welsh tend to be more Gaelic than the Scots who tend to be more so than the British.

    The North African is typical of a Spaniard as well, and while Galicia was never conquered by the Moors there could have been gene transfer, and they were in town from time to time. There were also ancient Phoenician and other African incursions, and some people in the north think they have Viking blood, due to the raids of the 8th-10th centuries. That's being studied, but the Vikings did not stick around to leave much DNA there.

    Looking at my results from myheritage:

    68.0% Iberian
    13.8% Scandinavian
    10.5% Northern And Western European
    6.6% North African
    1.1% Finish

    Here is an way to compare:

    The myHeritage side is a bit confusing. It seems like their reference population might be off. Then again, the Celts were in the part of the world shown as Germanic, and the Visigoths were Germanic, but only after they were Nordic. I could have some DNA from the Suevi or the Visigoths, and I can see how that might be difficult to untangle from even older Celtic DNA. These ancient populations largely overlapped.

    I will say that there are a couple of matches out of very very many that appear to be Scandanavian. The vast majority have Spanish or Lusophonic surnames, and are in Spain or the New World where Spaniards and Portuguese emigrated in the 20th century.

    My missing 3 x great grandpa would have only a 3% influence on my admixture, and it's possible that I have a similar missing relative I don't know about on my mom's side or two, so let's bump that up to 6%-10% overall. That's pretty close to the Eastern Euro estimates from ftDNA, but the nearly 15% of mystery DNA from myHeritage seems high.

    The Germanic vs Brittish Isles percentage is 17 vs 10.8. I could see a lot of crossover there with ancient Galician DNA, and I could see these things being hard to untangle.

    Let's say there is an Englishman in the reference population, and he says he is British as far as he knows, and he can go back all the way to 1200. That does not mean a large number of his ancestors at that point were not Germanic Angles or Gaelic Celts to a certain degree. Wouldn't such a problem make it harder to detect older DNA?

    In any case, I'm obviously having trouble reconciling the results, and I was wondering if anyone had any thoughts.
    Last edited by FuriousGeorge; 16 June 2017, 03:58 PM.

  • #2
    It's unfortunate that I can no longer edit my original post, because I've done a little research and come to understand that the term 'Celtic' is not really meaningful from a genetic standpoint.

    The inhabitants of NW Spain, Ireland, and Wales are in fact closely genetically related, but that is from migrations that first happened toward the end of the Ice Age.

    Later migrations of the people we call Celts apparently did more to spread culture than a unique shared genetic heritage.

    See also:

    The fact that the Basques are included show a connection to the pre-Basque Franco-Cantabrian culture, the difference being that the Basques remained more culturally and genetically isolated throughout subsequent migrations of Celtic, Roman, Germanic, and Moorish migrations.

    Understanding the results in that context, you would not expect reference groups from Western Europe to identify as Franco-Cantabrian, nor would that have much meaning so many milenia later.

    It seems what may be going on is a problem of resolution. The Germanic Visigoths were originally from Modern Day Sweden. England had several migrations from Nordic as well as Germanic peoples. The Suebi were Germanic, and there is evidence they interacted with Norse speaking people.

    I can see how non-Iberian and non-African DNA in me can be linked to just about anywhere, depending on the reference population.