Any thoughts on Admixture distributions re: the new update?

[neverJamToday]

The most recent update to ancestry compositions not only breaks things down quite a bit differently for some of the previously neglected regions but also seems to have done a number on groupings like British & Irish and French & German. French & German especially seems to have tightened up quite a bit. A lot of Broadly NW Euro is being reassigned and people's F&G percentages are going up and down. There's a German guy on Reddit right now with over 60% F&G, for example. My own fairly high F&G (51.7%) dropped down 6 points.

I wonder if the admixture distributions look like post-update, given what appears to be an increased resolution on these groups. I've previously proselytized on Reddit and the 23andMe forums explaining to people that no, less than 50% F&G doesn't mean you're not "100% German" but now I wonder how much that's true anymore...

[Angela]

The most recent update to ancestry compositions not only breaks things down quite a bit differently for some of the previously neglected regions but also seems to have done a number on groupings like British & Irish and French & German. French & German especially seems to have tightened up quite a bit. A lot of Broadly NW Euro is being reassigned and people's F&G percentages are going up and down. There's a German guy on Reddit right now with over 60% F&G, for example. My own fairly high F&G (51.7%) dropped down 6 points.

I wonder if the admixture distributions look like post-update, given what appears to be an increased resolution on these groups. I've previously proselytized on Reddit and the 23andMe forums explaining to people that no, less than 50% F&G doesn't mean you're not "100% German" but now I wonder how much that's true anymore...

There's a lot of shared ancestry between northern Europe and northwestern Europe, with lots of overlap between, for example, places like the Netherlands and Britain, Britain and Denmark and Norway etc.

National boundaries don't necessarily correspond with genetic boundaries. If people want to know from which countries their ancestors hailed in the last four to five hundred years, they should search their paper trail. These tests aren't going to reliably tell people those kinds of things, at least not people from certain areas in Europe.

The very fact that some companies can't tell the difference between the French and the Germans is a prime example. There is overlap between southwestern Germans and northeastern French. There's overlap between Eastern Germans and Eastern Europeans. There's overlap between northern Germans and Danes and Swedes.

You could say, well, for the French, use the Southwestern French sample, but that overlaps a bit with northern Iberians.

It's just not possible to tease it all apart and put a modern political label on it.

[kukumjacka]

In my case, 23andme confirmed what I had already discovered from my research of other genetic testings in the Balkans, well before 23andme and friends made DNA easily accessible.


I am now "highly likely" Greek, Albanian and Romanian. In addition, I am "likely" Bulgarian, Macedonian and Montenegroen. And furthermore, I am a "possible match" for Serbian, Croatian and Bosnian.

[Twilight]

There's a lot of shared ancestry between northern Europe and northwestern Europe, with lots of overlap between, for example, places like the Netherlands and Britain, Britain and Denmark and Norway etc.

National boundaries don't necessarily correspond with genetic boundaries. If people want to know from which countries their ancestors hailed in the last four to five hundred years, they should search their paper trail. These tests aren't going to reliably tell people those kinds of things, at least not people from certain areas in Europe.

The very fact that some companies can't tell the difference between the French and the Germans is a prime example. There is overlap between southwestern Germans and northeastern French. There's overlap between Eastern Germans and Eastern Europeans. There's overlap between northern Germans and Danes and Swedes.

You could say, well, for the French, use the Southwestern French sample, but that overlaps a bit with northern Iberians.

It's just not possible to tease it all apart and put a modern political label on it.

Do you think the new results is more accurate for detecting Iron Age Civilizations; Or was the old 23andme results better? :)

[Angela]

Do you think the new results is more accurate for detecting Iron Age Civilizations; Or was the old 23andme results better? :)

All European populations were formed by layers of civilizations.

What I do think is that as 23andme gets more samples it can be more accurate in its assignments to various areas.

However, the reality is that Italians, for example (from Italy), are in most cases 100% Italian, and have been probably for at least the last 500 years.

[Twilight]

All European populations were formed by layers of civilizations.

What I do think is that as 23andme gets more samples it can be more accurate in its assignments to various areas.

However, the reality is that Italians, for example (from Italy), are in most cases 100% Italian, and have been probably for at least the last 500 years.

Interesting, I’m curious now to see what the new updated 23andme map says. I did notice that my Germanic Ancestry plummeted and my Viking and Roman admixture disappeared. However I’m now curious to figure out why my Iberian admixture was elevated; I’m guessing my British Isles and French Ancestors might have been active seafarers to Iberia however with the Irish Milesian legends I got to wonder.
https://www.eupedia.com/europe/autos....shtml#23andMe

• British & Irish
  
  55.2%
  
  United Kingdom, Ireland
• 
  French & German
  
  19.9%
• Germany
• 
  Spanish & Portuguese
  
  3.5%
• 
  Eastern European
  
  2.8%
• 
  Finnish
  
  0.3%
• 
  Broadly Northwestern European
  
  16.1%
• 
  Broadly Southern European
  
  0.5%
• 
  Broadly European
  
  1.6%
• Sub-Saharan African
  
  0.1%
• 
  Congolese
  
  0.1%

[hrvclv]

The Beta version of 23andMe placed me right where I belong, geographically.

However, their breakdown assigned me 10.6% Irish British ancestry within the last 200 years, 10.6% Spanish Portuguese within the last 250 years, and 2.4% Italian within the last 300 years... when all my ancestors since 1550 (few of whom have escaped my research efforts) lived within 60 kms from my birthplace, and all had extremely local patronyms!

I guess their percentages give a ratio of genetic proximity with certain populations, at best, but certainly not "ancestry" proper, at least not within the time frames they take the risk of proposing.

[kukumjacka]

All European populations were formed by layers of civilizations.

What I do think is that as 23andme gets more samples it can be more accurate in its assignments to various areas.

However, the reality is that Italians, for example (from Italy), are in most cases 100% Italian, and have been probably for at least the last 500 years.

Our elevators at work are decorated with marble stones and a small sign notes the place of origin. One of them is Piana degli Albanesi in Sicily. Your reply brought that very spot to mind. The people living there are Albanians (Arberesh) who escaped the Ottoman occupation of Albania back in the late 1400s. To this day, they still speak the language, mantain traditions and traditional wear. In addition, the street names are in both Albanian (of that time) and Italian. Albanian is taught in schools and most people are bilingual.
Though I am sure through the centuries a great deal of interbreeding has occurred, I would find it quite interesting to see what the DNA of someone from that place would reveal of their origin. And this is right at that 500 year mark.

One very interesting thing I discovered some 15 years ago was that New Orleans has a fair amount of Arberesh Albanians from Italy who relocated there in the mid 1800s. And given how very diverse New Orleans is, it would be quite interesting to see a report from someone belonging to that particular group.

[Twilight]

The Beta version of 23andMe placed me right where I belong, geographically.

However, their breakdown assigned me 10.6% Irish British ancestry within the last 200 years, 10.6% Spanish Portuguese within the last 250 years, and 2.4% Italian within the last 300 years... when all my ancestors since 1550 (few of whom have escaped my research efforts) lived within 60 kms from my birthplace, and all had extremely local patronyms!

I guess their percentages give a ratio of genetic proximity with certain populations, at best, but certainly not "ancestry" proper, at least not within the time frames they take the risk of proposing.

Good news hrvclv, I brought this up to the 23andme team. It looks like they are going to investigate the admixtures. :)

Me: Hello there 23andme, I have an interesting boast that Eupedia site owner Maciamo have been making for a couple of years. Maciamo claims that 23andme can detect ancestry from Celts, Germanic tribes, Slavic and even Ancient Italian (Roman) Ancestry.
To prove such a boast, I was wondering if any Iron Age corpse have been tested on 23andme?
Would the Iron Age French have been 100% French/German or Lindow Man being of 100% British Isles?
Here is the link to where Maciamo boasts such a claim. Thanks in advanced :)
https://www.eupedia.com/europe/autos....shtml#23andMe
—————————————————————-
Hello,


Thank you for contacting the 23andMe Team. The analysis that we provide can only be performed using a saliva sample collected using the saliva kit that we include with your order. The saliva sample collected is ~2 ml. Unfortunately, we are unable to analyze any other type of sample.


Furthermore, our Ancestry Composition report assigns results based on your genetic similarity to our 45 populations worldwide. These populations are based on where your ancestors lived prior to the advent of intercontinental travel. This feature analyzes your autosomal DNA that you inherited from both of your parents and therefore reflects both your paternal and maternal ancestry.


For a full list of the reference populations you may see reflected in this report, visit: https://customercare.23andme.com/hc/articles/212169298


You can view more information about Ancestry Composition in the following help article: https://customercare.23andme.com/hc/articles/212861527


I hope you found this information helpful.


Best regards,


Kendall
The 23andMe Team
————————————————————
Me: That’s okay. Maybe there is another way to get the results without spit. Gedmatch has collected a variety of ancient samples. This page is outdated so there are undoubtedly more ancient ancestors whom have been DNA tested. Perhaps you can transfer their genes to 23andme. ;)
http://www.y-str.org/p/ancient-dna.html?m=1
—————————————————————
Hello,


Thank you for your reply. We appreciate your suggestions and have forwarded your comments to the appropriate team. As we strive to improve our service, your suggestions will be considered for incorporation into future updates.


Please feel free to let us know if you have any additional questions or suggestions.


Best regards,


Kendall
The 23andMe Team

[hrvclv]

Good news hrvclv, I brought this up to the 23andme team. It looks like they are going to investigate the admixtures. :)

Good effort! I'm not too sure their concern for what we write and how we feel will extend far beyond writing a polite, understanding answer, though. But my close relatives seem to think I am too much of a pessimist... :) so who knows?

[BertV]

I received a nearly triple dose of F&G after the changes. (21,3% to 63,1%) This could apply to more "natives".
With my papertrail it should be up-to 100% "G" when looking at a timeframe of < 500 years ago.

Would the Iron Age French have been 100% French/German or Lindow Man being of 100% British Isles?
Here is the link to where Maciamo boasts such a claim. Thanks in advanced :)

nope, I don't think it works this way.
The reference group is modern.
They probably would get a lot of broadly northwestern EU, unassigned and even more traces of surrounding population groups.