Ethnicity reports are only estimates
Here’s a quiz for DNA Sunday:
How many people are represented by the DNA results shown below?
DNA Result #1:
Great Britain 49%
Scandinavia 31%
Europe East 4%
Italy/Greece 4%
Europe West 4%
Iberian Peninsula 3%
European Jewish 2%
Ireland 1%
Africa North 1%
West Asia: Caucasus 1%
DNA Result #2:
West and Central Europe 87%
Scandinavia 11%
West Middle East 2%
DNA Result #3:
British & Irish 38.0%
French & German 27.1%
Scandinavian 4.2%
Broadly Northwestern European 24.2%
Southern European 1.4%
Eastern European 0.8%
Broadly European 4.1%
West African 0.2%
DNA Result #4:
Northwest European 66%
Mediterranean Islander 18%
Northeast European 12%
Southwestern European 4%
DNA Result #5:
Northern European 41%
Mediterranean 38%
Southwest Asian 21%
DNA Result #6:
North Sea 28.92%
Atlantic 29.51%
Baltic 11.42%
Eastern European 10.91%
West Mediterranean 8.07%
West Asian 5.5%
East Mediterranean 2.59%
Red Sea 1.65%
South Asian 0.64%
Northeast African 0.79%
DNA Result #7:
East European 12.97%
West European 49.33%
Mediterranean 23.65%
West Asian 10.75%
Northeast Asian 0.31%
Southeast Asian 0.11%
Southwest Asian 2.21%
African 0.65%
If you guessed that all of these results are for the same person — namely, The Legal Genealogist — give yourself a gold star.
The results, of course, are from different companies or services.
DNA Result #1 shows my current ethnicity estimate from AncestryDNA. DNA Result #2 is from Family Tree DNA. DNA Result #3 is from 23andMe. DNA Result #4 is from DNA.Land, and DNA Result #5 from National Geographic’s Geno 2.0 Project.
The last two both come from admixture calculators at GedMatch.com, with DNA Result #6 from the Eurogenes EUtest V2 K15 calculator and #7 from the Dodecad V3 calculator.
There have always been differences between the companies.1 But the differences aren’t getting any better over time; if anything, they’re getting worse.
Some of the differences, of course, are because of the different ways geographic areas are lumped together. At DNA.Land, for example, Northwest European includes “… Icelandic in Iceland; Norwegian in Norway and Orcadian in Orkney Islands.” That’s likely showing up as Scandinavian at AncestryDNA.
But the big reason why there are differences is because of what these admixture tests do: they take the DNA of living people — us, the test takers — and they compare it to the DNA of other living people — people whose parents and grandparents and, sometimes, even great grandparents all come from one geographic area. Then they try to extrapolate backwards into time. Nobody is out there running around, digging up 500- or 1,000-year-old bones, extracting DNA for us to compare our own DNA to.
So coming up with these percentages in these tests requires this fundamental assumption: that the DNA of the reference populations — those groups whose parents, grandparents, great grandparents and more all come from the same area — is likely to reflect what we might see if we could test the DNA of people who lived in that area hundreds and thousands of years ago.
In other words, these percentages are:
• estimates,
• estimates based on comparisons not to actual historical populations but rather to small groups of people living today, and
• estimates based purely on the statistical odds that those small groups tell us something meaningful about past populations.
These limitations are true of all of the testing companies. You can see from the above that my own results are — literally and figuratively — all over the map. I’m German with some companies, not German at all with another. Largely Scandinavian with one, only slightly Scandinavian with the others. A recent change in analysis at one company found my Germans, but in the process lost my British.
DNA testing for genealogical purposes is a wonderful tool. But people get disappointed when they see these percentages and they don’t match up to their own paper trail and don’t match up from company to company. And when they get disappointed, they may lose interest in genealogy or in DNA testing. And when they lose interest, we lose out on the paper trail information they might add to our mix.
The bottom line remains: We need to educate our friends and our families, our DNA cousins, to the limits of what these percentages can show — and to show them all the other things DNA testing really can help with.
Because it’s still not soup yet.2
And because we aren’t about to go digging up those old bones, it may never be soup.
SOURCES
- See e.g. Judy G. Russell, “Admixture: not soup yet,” The Legal Genealogist, posted 18 May 2014 (https://www.legalgenealogist.com/blog : accessed 15 Apr 2017). Also, ibid., “Making the best of what’s not so good,” posted 22 Feb 2015, and “Those percentages, if you must,” posted 14 Aug 2016. ↩
- For those too young to remember the reference, the Lipton Soup Company had a string of ads in the late 1960s and early 1970s. The mother in the ad would begin preparing Lipton soup, a child would ask, over and over, “is it soup yet?” and the mother would answer “not yet” over and over until finally she’d say, “It’s soup!” So “not soup yet” means “not finished” or “not ready.” ↩
I’ll trade you a 1% Eskimo for a 0.65% African.
SOLD!!! 🙂
Your ‘DNA Result #1’ only adds up to 82%. Did you forget something? Perhaps 18% of Europe West?
Thanks for sharing this comparison of results. It’s something I’ve been curious to see, but too cheap to pay for. All in all, I’m impressed with the “consistency”. A lot of it comes down to where they choose to draw the borders, or not draw borders (as in the case of ‘Result #3’. Clever how they have very specific categories (e.g. ‘French and German’, ‘Southern European’, and ‘Eastern European’); but then include non-specific categories to kind of fill in gaps (e.g. their “Broadly …” categories). In other words, it would seem to me that their ‘French and German’ category would be a portion of their ‘Broadly Northwestern European’ category, which likewise would be a portion of their ‘Broadly European’ category.
I am curious about the ‘Atlantic’ category of ‘Result #6’. What does that mean? After all, five continents border the Atlantic. Could this be for the lost city of Atlantis, maybe?
All in all, of the 5 company results, I’d say they are fairly consistent. There is only one company that looks odd and doesn’t seem to fit with the others. That would be the 21% Southwest Asian of ‘Result #5’.
Thanks for this post and the study it presents. I agree with your bottom line statement. DNA for ethnicity purposes is highly speculative.
Fixed, thanks. And Atlantic appears to be Scottish – Basque and a few other areas. (It’s hard to get a precise fix at the Eurogenes website.)
#2 has a big inconsistency with the others — no British Isles at all. I believe that company views West Central Europe and British Isles as two different regions, so with #1 and #3 both showing 30-50% British and Irish, #2 ought to have at least detected *some* British Isles inheritance.
It’s illegal to test DNA in France, creating a big empty space, and there seems to be no specific ethnic origins special to France either- a melting pot of different people started coming together for long periods of time there. That’s all I have to add…aside from my reports all varied to the extents used as examples as well. Taking them all and averaging, I still am Erin! My blood doesn’t feel any different now that I have confirmation of where it probably started around, even some surprise newer relatives that weren’t in my knowledge of heritage don’t change the way my blood feels, or morph my face into something it wasn’t before testing!
1. ” Nobody is out there running around, digging up 500- or 1,000-year-old bones, extracting DNA”…
2. “… for us to compare our own DNA to”
Actually, I think there are a whole bunch of people doing #1. It’s a growing field of research with new groupings of really old folks coming out daily it seems.
But no one seems to be doing #2 yet, Putting those old results into modern DNA databases. Which company will try first?
There are even some individual samples we can do one-to-one comparisons with at GedMatch, and Family Tree DNA does have the Ancient Origins feature, so… down the road maybe. But not today.
True. They are only just starting to see if usable DNA samples can be recovered from human remains found at sites where people once lived in ancient times (and some sites not so ancient, like Durham). Perhaps, one day, they will have enough samples to create a more reliable set of reference populations.
On the other hand, when you realize that some people traveled enormous distances (the Vikings, for example, raiding and trading everywhere from Russia to Greenland) testing ancient remains may only make the waters muddier than they are right now.
But, as with DNA for cousin matching, if we all decide to wait to have our own DNA tested until we are guaranteed accurate results, the data pool available for scientific analysis will remain too small to ever be able provide the kind of results we are waiting for.
I view it as an adventure, a voyage of discovery into an unknown realm. Like, Columbus, we may not know what we’ve found for many years to come, but if we don’t get on the ship and set sail, we’ll never find anything at all.
Yes, we need to take the results we may get today with a huge helping of doubt, but that shouldn’t stop us from getting joining the crew for the next wave of discovery.
There’s a website called My True Ancestry, which seems to be unreliable and has expensive, multi-tiered memberships. They also tell you that your free kit expires in 48 hours, then extends it.
There are a number of companies out there purporting to do this kind of analysis that I wouldn’t touch.
Judy,
I’d like to add a link to this post to the Hamilton Co. Genealogical Society website under the SWOHDNA Interest Group. With your permission, I previously linked your other articles on this topic.
Absolutely! And for a link alone (or a link with just a brief quote), you don’t need to bother asking. (But thanks for the courtesy!)
The problem is even with the digging up of old bones is how do you know where they really originated? Borders and ethnicity have been fluid throughout history.
Yep: dig up bones along Hadrian’s Wall and you could get Scots — or Romans.
I really got a kick out of familytreedna’s latest build! I previously was 44% Scandinavian (a lot of Viking naughtiness, I guess!)
Guess what? On their new one, I am now 0% Scandinavian! Sigh….
I lost all of my Siberian results and apparently become South American instead…
I still enjoy it immensely. I even like to come up with barely plausible reasons like a trace amount of Scandinavian comes from Viking raiders on the coasts of England (I come in 75-85% British, a good match to paper).
I’m tinkering with an alternative way to represent ancestry results here: http://www.impute.me/ethnicity/
…the goal I was going for was a more honest reporting, by showing all the reference-ethnicity samples on the same plot I thought it would be clear that even though you are right next to a, say, ‘english’ sample – then there’s also english samples that are way further away than, say, ‘finnish’. But all northern european samples are closer than, say, the african samples always.
Comments and feedback welcome
I love this — would love to see what it does with my data…
Yeah – try it. And let me know what you think – I’m still tinkering with the exact reporting mechanisms.
Unfortunately my Legal Genie … some people do not, and never will, understand the basic meaning of the word ‘estimate’ … they see it as a definitive result and we will never be able to satisfy their lofty expectations.
Knowledge is not power; the understanding and use of that knowledge is power … long live education 😉
Too true, I’m afraid… Sigh… 🙁
Hi Judy,
Thank you for your post.
I had a great laugh at the MyHeritage results for myself.
Cynically I decided that possibly they were developing a new marketing ploy when I noticed that they’d given me a fragment (0.8%) labelled Native American.
Which was a bit of a surprise given no other company does, and my known genealogy is 50/50 Scots/English – in present day at least.
My results from MyHeritage were also … um … less than overwhelming. 🙂
Nobody is digging up bones to compare us to? Wow. I guess you missed that entire section of Gedmatch. Strawman much?
The tiny handful of ancient DNA samples isn’t nearly enough to boil this soup. We have a very very long way to go on all of this.