And it’s still not soup
The Legal Genealogist was delighted earlier this week to discover that AncestryDNA has now, once again, found the Germans.
Not all of them, mind you, not even as many as there were the first time around, but at least some of them — and a lot more of them than the last time around.
Yes, we now are in the process of getting yet another set of those ethnicity estimate percentages that DNA testers love so much. It’s rolling out now for those who’ve tested with AncestryDNA and, when yours updates, you’ll see it as a option to preview an updated estimate:
Click on that and it will take you to the new percentages:
If memory serves, this is the third set of numbers from AncestryDNA and, for many people, each iteration has gotten at least a little bit better in terms of more closely approximating what we know of our ancestry on a paper-trail basis.1
For those of us with German ancestors, well, maybe not so much.
My ancestry is roughly 50% German and 50% British Isles-France mix. My father was born in Germany and I can trace his ancestry back there as far as there are church records. No Scandinavians that we can find — all solid Germans. My mother’s side is a colonial American mix of English, Scots-Irish and French Huguenot all blending in the North American melting pot since about 1670 or so.
The essential problem, it seems, is that there’s not a huge amount of difference between the ethnic signature of southern and southeastern England and that of northern Germany (and probably much of France and the low countries of The Netherlands, Belgium, etc., as well). And northern Germany may look a great deal like, say, Denmark. Or Sweden.
Which is why, in three iterations, my own percentages have gone like this:
| Version 1 | |
| Central European | 43% |
| British Isles | 35% |
| Scandinavian | 15% |
| Finnish/Volga-Ural | 7% |
| Version 2 | |
| Great Britain | 49% |
| Scandinavia | 31% |
| Trace region: Europe East | 4% |
| Trace region: Europe South | 4% |
| Trace region: Europe West | 4% |
| Trace region: Iberian Peninsula | 3% |
| Trace region: European Jewish | 2% |
| Trace region: Ireland/Scotland/Wales | 1% |
| Version 3 (as above) | |
| England and Wales | 66% |
| Germanic Europe | 19% |
| Norway | 7% |
| Sweden | 6% |
| Ireland and Scotland | 2% |
It’s better, yes. But it’s not soup yet.2
Remember what these numbers are: the only way to get these percentages is by comparing the test results of folks like you and me — alive today — to the test results of other people who are alive today (and not to the actual DNA of our ancient ancestors!!). Because of that, it’s all a numbers game based on one fundamental assumption: that people who live in an area today (say, modern Englishmen and -women) and who have all four of their grandparents born in that area are typical of the genetic signature of that population generations ago.
The numbers are terrific at the continental level: Europe versus Africa versus Asia.
But — particularly for those of us with western European and German ancestry — not soup yet below the continental level.
SOURCES
- See generally, for the change from version 1 to version 2, Judy G. Russell, “AncestryDNA begins rollout of update,” The Legal Genealogist, posted 13 Sep 2013 (https://www.legalgenealogist.com/blog : accessed 10 June 2018). Also, see ibid., “Those pesky percentages,” posted 27 Oct 2013. ↩
- The reference, for the youngsters in the crowd, is to the old Lipton Soup commercials 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 am glad to hear I may have some Germans in my ethnicity estimate. Mine hasn’t been updated yet. I’ve been blaming the Vikings for my Norwegian DNA, since I don’t have a paper trail to that region at all.
There are a couple of areas in England (Yorkshire), Ireland (Dublin?) and Scotland (Orkney and the Western Isles) where large numbers of Scandinavians settled and mixed with the native population many centuries ago, and modern studies have found substantial pockets of Norwedian dna persisting in those areas to this day. (In fact, Orkney was once actually part of the kingdom of Norway).
So, if your paper trail leads back to one of those areas of the British Isles, having a bit of Norwegian dna probably confirms your research is on the right track.
How many reference samples is your estimate drawing from? I’ve heard the new update includes “16,000 reference samples, 380 total regions” as opposed to the previous version of “3,000 reference samples, 363 regions”.
That’s what the preview says: “16,000 reference samples (13,000 more than before); 380 possible regions (17 new).” It doesn’t (yet) break that down any further.
It would be interesting to know what percentage of their additional reference samples are from non-European populations, and how many of the new regions are located in parts of the world that previously had little to no coverage. 97% of my estimate is exactly the same as before. The only difference is that instead of leaving 2-3% of my ethnicity unaccounted for, the addition of two new <1% regions on the fringes of Europe now brings the grand total reported up to about 99-100%
The African component has taken a step backward IMO. They have now removed Nigeria, opting for Cameroon and Southern Bantu as one region. Here is where it gets sloppy: they include Madagascar as a reference population in this region. This is a large problem as Madagascar’s founding population is Indonesian and yes, African populations are found there as well but isolated primarily to the western rim of the island. How they missed this I am uncertain.
The Asian components seem to be more granular as I now have “Phillipines” and “South Asian” (which speaks directly to my Malagasy ancestry).
My Native American is still of the broad “Americas” type and I do not expect that to change- ever! Lol.
As for my European component, the jury is still out. Both parents were fosters and my bio tree is quite versed but there are some small holes. There are still too high percentages for “broadly European” for my liking (30%). I am curious to know if the more ethnically diverse a person is the more difficult the read by the algorithms.
Now, for us African Americans we will NEVER have the granularity that the Europeans have, sadly. It will take a miracle to get enough Africans test not only atDNA but also M and Ydna. With all of the challenges on that continenent I do not see that happening any time soon.
Judy did you get an email saying your test was updated or did you just click on the Updates above the ethnicity bar graph? I have not received an email but simply clicked on the Updates last week (after seeing it and having heard about the update) and it appeared to update but nothing changed meaning all percentages are still the same as before.
My tests now say Up To Date above the ethnicity bar graph. But after clicking on the last test I manage, I noticed it actually gave a current version date at the bottom. Of my five tests I have these dates:
June 2014 [from my records original results received Jul 2014]
Feb 2018 [from my records original results received Aug 2017]
Feb 2018 [from my records original results received Aug 2017]
June 2014 [from my records original results received Jul 2014]
Sep 2014 [from my records original results received Oct 2014]
Apr 2015 [from my records original results received May 2015]
All version dates say 3,000 ref samples, 363 total regions.
Back in February the two Aug 2017 tests appeared to update but again the percentage numbers never changed from what was initially received (I realize that might happen) and apparently none of my earlier tests received an update at that time. (They also had no changes in the numbers.) Hopefully despite the different version dates the results are comparable to each other and not just the two sets with the same version dates (one set just happens to be twins now identified as identical.)
So if your version date says it uses 16,000 ref samples etc … I guess I’m still waiting for the update. I was interested to see what changes might happen and then disappointed again that there were none so maybe there is hope yet
If anyone else’s test showed no change check the version date, you might still be waiting too.
(a) No email, just a bar that reads “Preview an Updated Estimate” on the DNA origins stuff, and then the 16000 reference samples.
(b) It’s only my kit and none of the others I can access that have the update.
(c) It’s clearly a rollout and it’s going to take time.
That is what I saw too so I clicked but did not get the update. Hopefully the resulting message “Up To Date” eventually changes back to “Preview an Updated Estimate” again.
Is it true that there’s not much of a comparative base for French ancestry because the French are so concerned about privacy issues? I have a large “western Europe” chunk in my Ancestry ethnicity estimate, and my paper genealogy is strongly French Canadian.
It is true that not many people in France have tested. The same is true in many parts of the world.