Announcement

Collapse
No announcement yet.

A Problem With the FTDNA's Algorithm?

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • A Problem With the FTDNA's Algorithm?

    I took an AncestryDNA test back in January, and have sinced plugged that test into the likes of MyHeritage, DNA.Land, and Gedmatch. Most recently, I plugged it here into FTDNA and unlocked the MyOrgin results just a few days ago.

    Now, I know each company has their own regions, realize that regions can overlap, how people migrated, etc...so I'm rarely surprised when I get different totals for different sub-regions within regions/continents. So, for someone like myself who can trace most of his European ancestry back to England and Scotland (with a bit of German in there), I can understand how one test can show even much higher Scandinavian than the other. I can get why all of them are showing no "English" or low "Great Britain" given how mixed a population the British people are. I even have plausible theories on why I keep getting relatively high "Iberian" marks given no documented Spanish ancestors. I can get why given migrations, old and new, how in "Western" or "Northern" Europe you can get any mix of anything include Iberian, British Isles, Western Europe or Scandinavia.

    However, all have been fairly consistent in find that I'm about 35% European, and that generally over 80% of that is generally Nothwestern European (everything from France, Germany, UK all the way up to Sweden). This is confirmed through multiple family pedigrees going back hundreds of years and my DNA matches. So imagine my confusion knowing all of this, then, when I saw that while FTDNA showed the typical 35% European, that the single largest European region by far (15% of my total DNA and over 40% of my European DNA) was given to "Eastern European." Like, no other European region was even close. Even considering FTDNA's "Eastern Europe" region goes into eastern Germany where some other's companies end further east this makes no since. All other predictors have shown virtually no Eastern European DNA (<1%) or none at all; and unlike similarities within regions, Eastern and Western Europe are different enough to have much more clearly delineated borders than the sub-regions within Western Europe. This amount of "Eastern Europe" would have to mean very, very recent Eastern European ancestors, as in even as recently as over here in the U.S. where I'm from. I have none.

    To have consistently have been found to have little-to-no DNA from Eastern Europe, to have lines traced back hundreds of years in England and Scotland...there is just no way this result can be correct. Is this an error, does FTDNA has some unusually small or unrepresentative samples or something? Or simply a problem with FTDNA's algorithm? Like I said, I've been really tolerant of even wild variations within a region with these tests, but this makes NO sense to me.
    Last edited by WanderingHistory; 9th June 2018, 03:14 AM.

  • #2
    Join the club of odd ethnic origins results.

    The ethnic origins results of the DNA testing companies are highly questionable. You have to view it as a parlor game.

    By our four grandparents my brother and I are 25% Danish, 25% Irish and 50% Polish.

    Our ethnic origins breakdown is:

    Me Brother

    Europe East 63% 81%

    British Isles 9 12

    Scandinavia 0 7

    West Central 28 0

    Where did our Scandinavian go?

    Where is all that extra Europe East from?

    I can only guess that that Poland had a secret invasion of Denmark and Ireland that is lost to history.

    Comment


    • #3
      A lot of people have complained about getting too much Eastern European with the latest version of MyOrigins! My mother doesn't get any Eastern or Southeast Europe on any other ethnicity test or calculator but MyOrigins, which gives her plenty. Most of her ancestors came from a narrow band along both sides of the the Rhine River - Eastern France, Western Germany, Switzerland, Luxembourg, The Netherlands. That accounts for at least 75-80% of her ancestry, with the UK contributing the rest. I arrived at theses percentages through traditional genealogical research and they are much more precise than the various DNA companies' guesses because I know the names of the towns and villages in most cases.

      If you go back far enough to the migrations of the Franks, Huns, etc., I suppose she does have some Eastern European and so do most other Western Europeans. I don't think that's very meaningful to genealogists though.

      Comment


      • #4
        Thanks for this. Though, I think especially for the poster above with Polish ancestry this can be explained, and even west Germany is close enough that you can come up with theories.

        My example seems far more extreme knowing that I have next-to-no Eastern European ancestry, though. Going from 0% in one test to over 40% of my European ancestry in another is ridiculous knowing my background back for hundreds of years. Like I said, having a parent whose entire background is Eastern European and then getting higher than expected is within the normal range of these things. It's interesting, but not surprising.

        Comment


        • #5
          I agree with comments above about too much Eastern Europe being given out. My western czech heritage has resulted in 0% Western Europe even though I have many matches from Saxony , Bavaria and the Rhine region that other czech people don't share. I think the ftdna E Europe panel must include a fair few westerly czechs and poles who are carrying quite a bit of Germanic dna and as those SNP's are being read as E Europe , Germanic (West & Nth Germanic) people are picking up E Europe. Given that autosomal dna goes back reliably to probably the 1600's I would say based on what we know of germanic expansion into what is now Czech, Poland and even the baltic states during that time that it is more likely that the Eastern Europeans are carrying Western markers than the other way around.The slavic expansion into German territory happened probably from a 1000 yrs ago or more so would be an underlying strata in the Western German dataset resultingly and IMO is not the cause.

          Comment


          • #6
            You guys from the last two examples literally have significant ancestry from countries bordering Eastern Europe. Imagine the confusion for some with overwhelmingly British ancestry going back over 600 years (verifiably Anglo-Saxon) getting 40% of his European ancestry given over to "Eastern Europe." It makes no sense. My last full-blood European ancestor is a great-great grandfather. He'd have had to have basically been full Eastern European for me to be getting these ridiculous results.

            Comment


            • #7
              I sympathise with you that you've been scooped up wrongly into the Eastern European net! I expected to be at least 1/2 to 3/4 slavic markers and up to 1/2 germanic / celtic given both of my grandparents families stem for several hundred years from around 60 to 100kms from the Bavarian border. I am looking forward to a comprehensive improvement on the E Europe cluster when ftdna move to an origins 3.0. Come on Ftdna , get it together. IMO if Saxony and Thuringia and Prussia are included in the eastern european core it's always going to pull others from the north and west along with that. This area of Europe is quite complex and the dna will show it one day once the better granularity comes.

              Comment

              Working...
              X