What Dutch DNA Looks Like – 2020 Edition

Ethnicity estimates can vary between companies. They have different algorithms, different reference populations, and define different regions. To show you what Dutch DNA could look like with the various companies, here are my own results.

Based on my tree, I should be >99% Dutch, with a bit of German and French. I have some medieval lines from other countries in the 1200s and before, but the chances of me having inherited any of their DNA is minimal. My father’s side is from Gelderland, near the German border, and my mother’s from Noord-Brabant and Zeeland, near the Belgian border. So how did the companies do?

Companies fine-tune their predictions over time. This is the third article I’ve written about this. I left the previous ones online, so you can compare the current estimates to the ones in 2016 and 2018.

Ancestry

Ancestry ethnicity estimates

Ancestry thinks I’m 53% England, Wales & Northwestern Europe, 45% Germanic Europe, and 2% Swedish. The overview map shows the first category located in England and Wales, but if I click on that category, is shows a lighter area that includes the southern part of the Netherlands where my mother’s family is from. So although I find it weird for that part of the Netherlands to be lumped in with England and Wales, I can’t say it’s incorrect.

The Germanic European, which Ancestry thinks belongs to the genetic community Netherlands, is spot on. I don’t know where the 2% Swedish DNA came from, but that might be noise.

Interestingly, my mom has the same roughly 50/50 split between English/Wales/NW Europe and Germanic Europe, so it’s not like all of her DNA was responsible for one predicted ethnicity and my father’s ancestors on the German border were responsible for the other.

There is a big difference since 2018, when Ancestry had me as 94% Western European, and low confidence regions Great Britain (2%), Iberian Peninsula (2%), Ireland/Scotland/Wales (1%) and Scandinavia (<1%). The Iberian disappeared completely, but was marginal to begin with.

23andMe

23andMe predicts me to be 58.1% French & German, 22.9% British & Irish, 4.2% Scandinavian, 14.3% Broadly Northwestern European, 0.1% Broadly European, and detected trace ancestry of 0.3% Anatolian, 0.1% Broadly Central Asian, Northern Indian & Pakistani.

The French & German has gone up since 2018, which is good since according to my research that should be 100%.

If I click through to the regions for French & German, it identified Gelderland and Noord-Brabant as the top two positions. That is consistent with my tree. Color me impressed! I hardly have any known ancestors from North and South Holland, so I do not understand why they appear so high. Zeeland, no. 5 according to 23andMe, is in third place in my tree.

It also predicts I have ancestors from Flanders (more likely) and Walloon (less likely) in Belgium in the past 200 years, which does not fit my tree. But since my grandparents were born a couple of miles from the Belgian border, I think the Flemish prediction is not that far off.

For Germany, it predicts my ancestors likely came from North Rhine-Westphalia. I do indeed have an ancestor, Arend Kastein, who was born in Westphalia in 1817, just over 200 years ago. Several of his ancestors were from Westphalia too, as well as some other ancestors of mine further back. I don’t know of any ancestors from other parts of Germany, but these are less likely according to 23and Me.

All in all, 23andMe’s regional predictions are pretty impressive, in my opinion. It’s too bad about the large amounts of unexplained British.

FamilyTreeDNA

FamilyTreeDNA estimates me to be 89% British Isles, 4% Southeast Europe, 2% Asia Minor, with trace results <2% Central Asia, <1% South America, and <2% West Middle East. The results for the are the same as in 2018.

Like in the previous years, FamilyTreeDNA completely missed my West and Central European DNA, which it does detect in my mother (61%) and paternal uncle (59%). They both also have significant amounts of Scandinavian (15% and 29%, respectively), where I have none. This clearly shows a problem with their algorithm. My tree is >99% Dutch, which should be in the West and Central European ethnicity, so it is hard to understand why they fail to pick up any when they do find it as the largest ethnicity for my mother and paternal uncle.

MyHeritage

MyHeritage predicts me to be 64.2% North and West European, 32.8% Irish, Scottish, and Welsh, 1.8% Askenazi Jewish, and 1.2% Middle Eastern. The results have not changed since 2018. I don’t know of any Irish, Welsh, Jewish, and Middle Eastern ancestors. I have one Scottish mercenary soldier in my tree who arrived in the Netherlands in the early 1600s, but he alone cannot account for being estimated to be almost 1/3 Irish, Scottish, or Welsh. I’d love to be, but I’m just not.

LivingDNA

Living DNA estimates me to be 41.5% Great Britain and Ireland, 36.6% Northwest Germanic, 12.4% Scandinavia, 6.7% South Germanic, and 2.8% Basque. The Great Britain is further split up into predictions for East Anglia (16.5%), Central England (10.6%), South Central England (6.5%), Cornwall (3.5%), Southeast England (3.4%), and South Yorkshire (1.1%).

In terms of their categories, I should be >99% Northwest Germanic, so they’re pretty far from the mark. The 41.5% Great Britain and Ireland is way too high, and the subdivision into regions within England makes no sense. However, only 41.5% British is a huge improvement over the 2018 estimates, which had me at 94.4% British/Irish.

Comparison

2016 2018 2020
Ancestry 94% Western European
2% Great Britain
2% Iberian Peninsula
1% Ireland/Scotland/Wales
<1% Scandinavia
94% Western European
2% Great Britain
2% Iberian Peninsula
1% Ireland/Scotland/Wales
<1% Scandinavia
53% England, Wales & Northwestern Europe
45% Germanic Europe (Netherlands)
2% Sweden
23andMe 37.7% French & German
28.2% British & Irish
6.5% Scandinavian
25.1% Broadly NW European
1.4% Eastern European
1.2% Broadly European
39.3% French & German (Netherlands)
28.9% British & Irish (United Kingdom)
6.8% Scandinavian
1.4% Eastern Europe
22.6% Broadly Northwestern European
1.1% Broadly European
58.1% French & German
22.9% British & Irish
4.2% Scandinavian
14.3% Broadly Northwestern European
0.1% Broadly European
FamilyTreeDNA 40% British Isles
27% Scandinavia
22% Southern Europe
6% Eastern Europe
3% Central/South Asian
2% Middle Eastern
89% British Isles
4% Southern European
2% Asia Minor
89% British Isles
4% Southeast Europe
2% Asia Minor
MyHeritage not tested 64.2% North and West European
32.8% Irish, Scottish, and Welsh
1.8% Ashkenazi Jewish
1.2% Middle Eastern
64.2% North and West European
32.8% Irish, Scottish, and Welsh
1.8% Askenazi Jewish
1.2% Middle Eastern
LivingDNA not tested 94.4% British/Irish
1.1% North-Western European
2.5% North Turkish
1.9% North-Western Caucuses
36.6% Northwest Germanic
12.4% Scandinavia
6.7% South Germanic
41.5% Great Britain and Ireland
2.8% Basque

We can see from this table that all companies except FamilyTreeDNA and MyHeritage updated their estimates since 2018. This shows that it pays to go back to your results, since they change over time.

In general, I am more pleased with my results now than I was in 2018. I do feel like Ancestry is cheating a bit by lumping North-Western Europe in with England and Wales, which the other companies consider to be separate ethnicities. I think it’s impressive that Ancestry predicted I belong to the “Netherlands” genetic community, and that 23andMe not only picked the Netherlands as the most likely country of origin for my ancestors, but even pegged Gelderland and Noord-Brabant as my top provinces.

What these results show more than anything is that we should not treat them as gospel. Even people whose ancestors pretty much all come from the same country don’t get uniform results. We shouldn’t freak out if the ethnicity predictions don’t match our trees.

At the continent level, the predictions were solid. The companies all pegged me as 100% or near 100% European, which matches my tree. Below that, they all placed me in the North-Western part of Europe, though some thought I was more British than I am. At the lower levels, they become increasingly unreliable, which is easy to understand if we look at the migration history of Europe.

If I have to pick a favorite, it would have to be 23andMe, because they got the highest prediction of North-Western European or Germanic DNA (not counting Ancestry’s combined category) and got my top two provinces right. But Ancestry and MyHeritage also did well. LivingDNA suggesting areas in England that have no relation to my tree makes them less believable in my opinion, though they have improved since 2018. I think FamilyTreeDNA knocked themselves out of the competition by not detecting any of my West and Central European DNA. I look forward to seeing how these results improve over time.


Note to lecturers:
Feel free to use the screenshots for your own presentations about genetic genealogy, if you credit me as the contributor. I’d appreciate if you drop me a line if you use the screenshots since I like hearing how they’re used.
About Yvette Hoitink

Yvette Hoitink is a professional genealogist in the Netherlands. She has been doing genealogy for almost 25 years. Her expertise is helping people from across the world find their ancestors in the Netherlands. Read about Yvette's professional genealogy services.

Comments

  1. Karin (Rozema) Kuiper says

    Hello, Yvonne. Thanks for sharing this. My grandparents all lived in Groningen for centuries before coming to the U.S. My earliest Ancestry results showed me firmly rooted in northern Netherlands, Norway, and Sweden. I am confused that more recent analysis shows a connection with the British Isles. It seems tribes and Vikings traveled south to Netherlands and Britain, carrying DNA and accounting for the same genes among the people. Am I wrong about that?

    • Old Frisian and Old English are very similar. Too much emphasis has been placed on Anglo-Saxon migration to England and too little on Frisian sea-faring merchants.
      You might enjoy reading The Edge of the World by Michael Bye.

  2. This is fascinating – thanks for sharing your analysis, Yvette! I’m tempted now to test (and have my parents test) with 23andMe, to compare with their results from the other companies.

  3. Are data available on the U.S. which would show Dutch genomic content which could be matched to provinces in the Netherlands?

  4. From what I understand, the Netherlands were under Spanish rule for quite some time and many modern day Dutch have Spanish ancestry (I believe I’ve read one out of five, at least? but I can’t remember the exact statistic). Perhaps the Middle Eastern & Jewish could be remnants of that era of rule — possibly a Spanish-Dutch ancestor with strong/mostly Middle Eastern/Jewish heritage?

    • Kaniel Outis says

      After 1492, when Sephardi Jews were expelled from Spain, Amsterdam became a Sephardi cultural hub. Many Sephardi intermarried, hence there are non Jewish Dutch with Jewish ancestry. There are a small amount of Dutch males (who have been sampled) who have the Y-Haplogroup r1a-z93, and this can be traced back to Ashkenazi Levite ancestry.

  5. neil devers says

    Thanks for your interesting comments. I have tested with Ancestry, FTDNA, My Heritage and Living DNA. Broadly speaking my experience matches yours. Some of Ancestry is spot on with what I know, others pretty vague. FTDNA I have found the least accurate. Most of Living DNA matches with what I know (but not all). My Heritage OK but pretty vague. It seems Sweden is pretty common – Although I don’t know anyone from there. Thanks, Neil

  6. Geoffrey Pugh says

    Earlier this year I took a test with Ancestry which gave a result of 84% England, Wales & N.W. Europe and 3% France. Since my mother and grandmother were both born in France (which I have evidence to prove this), I decided to take another test with My Heritage that gave a result of 76.8% North and Western European!
    I have sent four e-mails to Ancestry and finally spending over an hour on a telephone call to the company querying their results. All to no avail. I suggested they refund me the fee or else provide another test kit free of charge.
    In view of my experience, I would not recommend anyone to waste their money on taking a test. For a company of their standing I am staggered at the response I received to my complaint.

  7. Kaniel Outis says

    Autosomal DNA represents where populations cluster, as opposed to where an individual’s relatives were born.

    Living DNA use PCA charts to determine results. Their regions are the most accurate when it comes to autosomal clusters. Dutch people cluster close to southeast English, hence they may get results that suggest they are English. Their results have become much better since their 2020 updates, as they have added new regions, which means they no longer overestimate their British results.

    Given that you are Dutch, it doesn’t surprise me that you score Northwest Germanic and English, as PCA charts show that this is where Dutch cluster. These results don’t make you any less Dutch. They just tell you that you cluster, primarily, with East English, Dutch, and northwest Germans.

    I am Eastern European yet I score higher Southern European and Near Eastern results, than Eastern European ones, as my ethnicity is Ashkenazi. Ashkenazi Jews may be from countries like Poland, and Ukraine, yet on an autosomal level they do not cluster with East Europeans.

  8. carolyn Adams says

    Very interesting article. Thank you. I always work on the assumption that I do not have a caste iron understanding of my ancestors although i have been researching them for 7 years. For example there is no guarantee that a fathers name on a birth certificate is correct. Further I have no knowledge of a Scandinavian ancestor but get many distant matches from there on both Living DNA and Familytree DNA.

  9. Like you, my husband with Frisian parents had quite a bit of British DNA. Myself, I’m from a British background and had some western European DNA. Both of us were surprised, but the more research I
    do on the history, the more I am not surprised.

    Flemish weavers were in high demand in England and Scotland. Huguenots fled western Europe for Scotland and England, and the Netherlands was Scotlands greatest trade partner at one time. Both countries also fished the same waters and met on the northern islands (exchanging knitting patterns amongst themselves and with Scandinavians as well)

    I read in a book that at one point one of the little NE Scotland fishing villages had so many Dutch coins in the church collection they had to put a stop to it.

    I know people are intensely proud of a certain heritage, but we have to open our eyes up to the fact that the sea ways were the mode of transportation and these people did travel.

    Of course results also depend on the people who test. If more British people than Dutch people are using a particular testing site, then that is going to skew results as well.

Leave a Reply to carolyn Adams Cancel reply

*