PDA

View Full Version : Random Assignments for Central Europeans of German Descent



Riverman
07-05-20, 00:07
23andme improved a lot over the years, looking at where they started and how good they are now. But for Central Europeans there is still a lot of room for improvement, especially below the major divisions of North Western European and Southern European. I selected a fairly homogeneous group of Southern Germans and Western Austrians for a comparison to prove my previously made observations and the result is quite clear.
In this group of people the NWE and SE stays pretty constant overall (single exception for SE), what changes is primarily the assignment to the subdivisions. To give an example from a homogeneous 5 sample group for the ranges:
NWE: 81,1-83,9 (= 2,8)
F&G: 58,8-68,1 (= 9,3)
B&I: 2,1-5,2 (= 3,1)
Scandi: 0-1,5 (= 1,5)
BNWE: 8,3-18,9 (= 10,6)

Notice the only slight difference in general NWE and the large discrepancies between the subdivisions.

So everything considered, it is pretty obvious that NWE largely equates German in Central Europe, especially considering that 23andme concentrates on more recent regional ancestry anyway. The assignment to the subdivisions is generally not that far off, but takes just away from F&G, probably because a large portion of the calculation was done with mixed American citizens (?) and the difficulties in differentiating between the NW components in general.

The numbers for B&I, Scandi and especially BNWE increase in direct proportion to unknown and bad sampling. People from undersampled German subpopulations get much higher BNWE. Like a German with more Eastern German ancestry and a lower general NWE below 70 percent gets almost 30 percent for the remaining NW, especially 17 percent for BNWE. This is because the algorithm is still able to recognise the German ancestry as NWE, but because of a lack of proper sampling assigns it wrongly or refuses assignment (BNWE, still the better option). Sometimes, the German ancestral component even ends up in "Broadly European" because the NWE signature is not the clear cut samples from 23andme.

For the Southern European part, which is clearly leaning towards Italian in the homogeneous Southern German-Western Austrian sample, the same is somewhat true, but not as much for the subdivisions. Sometimes parts of the SE component just end up in "Broadly European" too, even though the algorithm should have recognised it better. Generally "broadly" seem to appear the worse the sampling is and the more mixed the individual is, because of the large segment assignment 23andme applies.

Its however amazing how consistent the general NWE ancestry is in this homogeneous group, that's at least a fairly robust result. But like I wrote in the title, improvement especially through better sampling is definitely possible.

I guess the same patterns being largely true for other ethnic groups from Europe? Or are the subdivisions better and more consistent somewhere?

Regio X
07-05-20, 01:28
23andme improved a lot over the years, looking at where they started and how good they are now. But for Central Europeans there is still a lot of room for improvement, especially below the major divisions of North Western European and Southern European. I selected a fairly homogeneous group of Southern Germans and Western Austrians for a comparison to prove my previously made observations and the result is quite clear.
In this group of people the NWE and SE stays pretty constant overall (single exception for SE), what changes is primarily the assignment to the subdivisions. To give an example from a homogeneous 5 sample group for the ranges:
NWE: 81,1-83,9 (= 2,8)
F&G: 58,8-68,1 (= 9,3)
B&I: 2,1-5,2 (= 3,1)
Scandi: 0-1,5 (= 1,5)
BNWE: 8,3-18,9 (= 10,6)

Notice the only slight difference in general NWE and the large discrepancies between the subdivisions.

So everything considered, it is pretty obvious that NWE largely equates German in Central Europe, especially considering that 23andme concentrates on more recent regional ancestry anyway. The assignment to the subdivisions is generally not that far off, but takes just away from F&G, probably because a large portion of the calculation was done with mixed American citizens (?) and the difficulties in differentiating between the NW components in general.

The numbers for B&I, Scandi and especially BNWE increase in direct proportion to unknown and bad sampling. People from undersampled German subpopulations get much higher BNWE. Like a German with more Eastern German ancestry and a lower general NWE below 70 percent gets almost 30 percent for the remaining NW, especially 17 percent for BNWE. This is because the algorithm is still able to recognise the German ancestry as NWE, but because of a lack of proper sampling assigns it wrongly or refuses assignment (BNWE, still the better option). Sometimes, the German ancestral component even ends up in "Broadly European" because the NWE signature is not the clear cut samples from 23andme.

For the Southern European part, which is clearly leaning towards Italian in the homogeneous Southern German-Western Austrian sample, the same is somewhat true, but not as much for the subdivisions. Sometimes parts of the SE component just end up in "Broadly European" too, even though the algorithm should have recognised it better. Generally "broadly" seem to appear the worse the sampling is and the more mixed the individual is, because of the large segment assignment 23andme applies.

Its however amazing how consistent the general NWE ancestry is in this homogeneous group, that's at least a fairly robust result. But like I wrote in the title, improvement especially through better sampling is definitely possible.

I guess the same patterns being largely true for other ethnic groups from Europe? Or are the subdivisions better and more consistent somewhere?I guess most of testees are from Central and North Europe, which may explain why most of actual Central and North Europeans consistently get a huge amount of NW Euro. In other words, perhaps 23andMe chose to be more accurate for N. Europeans. I don't know... Anyway, it all depends on the reference samples chosen by 23andMe for each cluster, and the following map (from 2017), showing where each component would peak, may provide clues on what the references are.

12047

My mother, for example, still gets more than 40% of NW Euro, which wouldn't make much sense for a N. Italian. If they "corrected" it and made her get more S. Euro %, they would possibly do it at expense of another cluster. That's the problem. Anyway, those are not actual NW Euro % in her case. It must be shared ancestry, especially if the reference samples for F&G come from around the Alps.
Anyway, Switzerland seems a decent choice for F&G, since Alps works as a genetic barrier, and a more Northern reference could mess up with Scandinavian cluster, for example. Complicated!

We discussed something about it in the following thread. Hope it helps some way:
https://www.eupedia.com/forum/threads/38231-23andMe-s-methodology-artificially-purifies-your-ancestry-results

Not sure my answer has to do with your post. Later I'll check it again, closer, since I'm a bit hurry now. :)

Riverman
07-05-20, 02:12
Thanks, I knew the sampling issue of 23andme and think too it has a lot to do with the "weight" of the samples and customers. Those which are well represented in both categories get the recognition bonus. Those which are undersampled and underrepresented don't. At least that's a very simplified explanation of what's probably much, much more sophisticated and complicated ;)

However, I made the additional observation that somewhat more Northern, Central-Western Germans might get even higher numbers for G&F as well as NW. Like about 95 for NWE and 75-80 for F&G. Yet if the Swiss sample is still the gold standard for F&G, that's strange, no? I have to confess however, that I have not seen too many Swiss German results.

Here is one example I could find:
https://imgur.com/a/XorTq

https://www.reddit.com/r/23andme/comments/87kfgl/my_swiss_results/

The questions she had show how serious people can take results of let's say British & Irish, as if this proves recent ancestry, its just a model calculation from a company. The results are from before the most recent updates though, so probably might look different now (?).

An issue 23andme always had, and it improved on that, is not considering regional ethnic variation enough. Like sometimes they seem to include regional samples from a very distinct ethnicity which is much closer to a different people and region into the regional sampling. They did improve on that, but I think they still include outliers? Probably even deliberately because they focus on regional rather than ethnic differences? Yet this distorts the bigger picture, because then they get more DNA segments they can't assign, because they are present in too many places. The solution is to assign segments to neighbouring ones and producing larger ones for assignment, isn't it? Yet that's get harder the more mixed a person is, with wrong positives and negatives. But there is no easy solution to it probably, especially if the genetic distance between two populations wasn't that big and largely overlaps anyway.

Regio X
07-05-20, 02:54
@Riverman
IIRC, they say LivingDNA is really good only for British, for example. It may be complex to calculate recent ancestry, as you suggested, and they perhaps made a "comercial" choice.
Ancestry Composition seems the best option available anyway.
Concerning the results of that Swiss girl, she says 5 great-grandparents were from Switzerland. The others would be from Italy and Austria, if I got it right. Either way, there're certainly regional differences in Switzerland, as between Italian and German speaking areas.
Notice that the map doesn't show averages. It shows the areas where the author found the highest % for certain cluster, but significant variations in zome of these areas may be possible. Finally, it's three years old, so something could have changed since 2017, yes.

Riverman
07-05-20, 11:46
By the way, I think the NWE in Italians is legitimate. Because of the Northern influx. The percentage seems to correlate with Germanic influences too, like South Tyrol and Lombardy having higher numbers than regions without German/Germanic settlement. Did anyone take an ancient sample, like of a pre-Imperial Roman for a 23andme test? Would be fun results probably ;)

Regio X
07-05-20, 13:13
By the way, I think the NWE in Italians is legitimate. Because of the Northern influx. The percentage seems to correlate with Germanic influences too, like South Tyrol and Lombardy having higher numbers than regions without German/Germanic settlement. Did anyone take an ancient sample, like of a pre-Imperial Roman for a 23andme test? Would be fun results probably ;)My mom is Venetian in ancestry.
MyHeritage's calculator and myOrigins are certainly worst than Ancestry Composition. I didn't test at Ancestry.com, but apparently its calculator doesn't score that well as Ancestry Composition in ISOGG.
https://isogg.org/wiki/Autosomal_DNA_testing_comparison_chart

That said, I don't think my mom's NW % de per si makes much sense, because 23andMe focus on recent ancestry, but also because Y-DNAs in N. Italy don't seem to support such impact in Autosomal by Lombards, Goths etc. If my memory serves me, they came in relatively low numbers.
That's why I think most of those % must be some shared (older) ancestry with Central E., rather than actual ancestry or "Germanic" ancestry.

Riverman
07-05-20, 14:03
My mom is Venetian in ancestry.
MyHeritage's calculator and myOrigins are certainly worst than Ancestry Composition. I didn't test at Ancestry.com, but apparently its calculator doesn't score that well as Ancestry Composition in ISOGG.
https://isogg.org/wiki/Autosomal_DNA_testing_comparison_chart
That said, I don't think my mom's NW % de per si makes much sense because 23andMe focus on recent ancestry, but also because Y-DNAs in N. Italy don't seem to support such impact in Autosomal by Lombards, Goths etc. If my memory serves me, they came in relatively low numbers.
That's why I think most of those % must be some shared (older) ancestry with Central E., rather than actual ancestry or "Germanic" ancestry.

Couldn't it be that its both? But especially Celtic and Germanic? The study on Italians measured a noticeable shift in the direction of the German people from late Antiquity to Medieval times. I think they were onto something and that this is real and not just sample bias. Because yes, the rural Italics survived better than the mixed urban population with its Eastern Mediterranean shift, but that doesn't explain the whole trend.

If its something older, the decrease shouldn't be as pronounced. I don't know on which date (before or after the latest update) this map is based upon, but the F&G decrease is almost abrupt from North to South according to this map from Maciamo:
https://www.eupedia.com/europe/autosomal_maps_dodecad.shtml#French_German

Old Italian variation is not as good as explaining this as later Celtic and Germanic settlement on top of it. BNWE basically repeats that pattern, with the more clearly German parts having a subtraction of their F&G (South Tyrol in particular).

The pattern is the same in regions like Finland by the way, where you just have to substitute F&G with Scandinavian.

Regio X
07-05-20, 15:48
Couldn't it be that its both? But especially Celtic and Germanic? The study on Italians measured a noticeable shift in the direction of the German people from late Antiquity to Medieval times. I think they were onto something and that this is real and not just sample bias. Because yes, the rural Italics survived better than the mixed urban population with its Eastern Mediterranean shift, but that doesn't explain the whole trend.

If its something older, the decrease shouldn't be as pronounced. I don't know on which date (before or after the latest update) this map is based upon, but the F&G decrease is almost abrupt from North to South according to this map from Maciamo:
https://www.eupedia.com/europe/autosomal_maps_dodecad.shtml#French_German

Old Italian variation is not as good as explaining this as later Celtic and Germanic settlement on top of it. BNWE basically repeats that pattern, with the more clearly German parts having a subtraction of their F&G (South Tyrol in particular).

The pattern is the same in regions like Finland by the way, where you just have to substitute F&G with Scandinavian.Yeah, perhaps influences of groups such Celts and Rhaetians could help to explain it, but these are folks a bit older. Indeed, the NE is rich in R-L2 and relatively rich in G-L497. Plus, first Italics possibly had some resemblance to the formers, I believe, and if the latter were close to Etruscans also in Autosomal, then they would not have been that different from Italics either. That Roman soldier FN2 from 300 AD found in Bavaria was likely too late for a Rhaetian proper, but he could have been sort of a "Rhaetian relic", je je, since he was somewhat shifted to modern Iberians - just as Etruscans - and S. French IIRC, at the same time he carried a notorious "Alpine" Y marker. Plus, mountains are more prone to genetic continuity. So who knows!
"The Raeti tribes quickly became loyal subjects of the empire and contributed disproportionate numbers of recruits to the imperial Roman army's auxiliary corps."
(...)
"Further support for the hypothesis that the northern Raeti tribes converted to Celtic speech before the Roman imperial era is provided by the distribution of 'Raetian' inscriptions. These have been found almost exclusively in northeastern Italy: South Tyrol, Trentino, and the Veneto region. None have been found in Switzerland, the other core Raeti region. The Raetic inscriptions indicate that 'Raetian' survived as late as the 3rd century AD, suggesting that the Raeti tribes in this region at least may not have converted to Celtic speech." (Wikipedia)
Anyway, all of them must have been somewhat rich in EEF ancestry as well, even Celts (as per the shif in Iron Age UK). It's possibly someway involved in this "shared ancestry", naturally, while modern Italian cluster must also include other components, in turn not that strong in North Italy.

Finally, later movements probably helped to shape modern NE Italians, and we could include Germanics', sure, as well as movements from the South.
As for Germanics specifically, I said most part of that % in my mom could not be assigned to them, however, I agree that some of it could be. Not sure how much, but Y-DNA may provide clues in this regard. R-U106 must account for ~5% around Treviso province, and I1 frequency would be three times bigger in there. So perhaps something close to 20% around Treviso province specifically (where most of her ancestors were born). Now, if we think that the impact over Autosomal might have been substantially lower than that...

Thanks for the Eupedia's map on F&G. Notice that 42.2% is actually her NW, not her F&G, which is 34.8%. My father, on the other hand, scores 34.6 in the former and 24.5 in the latter, plus 1.6 East Euro, which is curious (my mom doesn't score any), also because he's basically 75% Venetian in ancestry, not fully. The other 25% come from far East Lombardy.

Riverman
07-05-20, 16:03
Notice that 42.2% is actually her NW, not her F&G, which is 34.8%. My father, on the other hand, scores 34.6 in the former and 24.5 in the latter, plus 1.6 East Euro, which is curious (my mom doesn't score any), also because he's basically 75% Venetian in ancestry, not fully. The other 25% come from far East Lombardy.

The fun thing about your parents results is that your mother has more F&G than many actual Germans! But her NW is not higher. So far I never saw a full German (by recent ancestry) with a lower value than 50 percent for NWE and in Germans that's decisive. So it must be also about the Celtic and French-like ancestry of the Northern Italians which play in, since F&G is not exclusively German, at least for the West and South. The uniparentals are a tricky thing, because one has to look at both sexes and consider the variation which present among the Germanic incomers already. They were not necessarily all that typical Northern European to begin with, but already more like modern Southern Germans.
Modern Southern Germans and French are also the most likely source for more modern admixture in the time of the Holy Roman Empire.

If you subtract the Southern Papal states, the border of the Holy Roman Empire in Italy is almost the same as the one for NWE admixture in Italians:
https://en.wikipedia.org/wiki/Holy_Roman_Empire#/media/File:Holy_Roman_Empire_at_its_territorial_apex_(pe r_consensus).svg

Might be more coincidental and without too much of a relevance, but who knows what comes up next.

Regio X
07-05-20, 18:40
The fun thing about your parents results is that your mother has more F&G than many actual Germans! But her NW is not higher. So far I never saw a full German (by recent ancestry) with a lower value than 50 percent for NWE and in Germans that's decisive. So it must be also about the Celtic and French-like ancestry of the Northern Italians which play in, since F&G is not exclusively German, at least for the West and South. The uniparentals are a tricky thing, because one has to look at both sexes and consider the variation which present among the Germanic incomers already. They were not necessarily all that typical Northern European to begin with, but already more like modern Southern Germans.
Modern Southern Germans and French are also the most likely source for more modern admixture in the time of the Holy Roman Empire.

If you subtract the Southern Papal states, the border of the Holy Roman Empire in Italy is almost the same as the one for NWE admixture in Italians:
https://en.wikipedia.org/wiki/Holy_Roman_Empire#/media/File:Holy_Roman_Empire_at_its_territorial_apex_(pe r_consensus).svg

Might be more coincidental and without too much of a relevance, but who knows what comes up next.Well noticed. If my parents had that "Germanic" influence, we should perhaps expect, say, some Scandinavian %. I believe Germans score some, and Venetians get only traces, if any. I used to score 0.7 or something, but now I score 0. Ideally, we should also know what those Germanics looked like compared to modern Germans, after all, we're talking on an old influence. See how Italy changed along the time. Of course, the changes may have been higher in one area than in another; still...
As you suggested, it'd be also interesting to see how Italics themselves would score. My guess is that they were not "that" different from us, in part "accidentally", given those later influences in Veneto from both South and North. For a similar reason I think my great proximity to the "Illyrian" at MTA is in part "accidental". I'm certainly not an Illyrian relic. :)

Still concerning Germans, it may be true they don't score amazingly high in F&G, however, they score more than 90% in NW Euro, right? Very few S. Euro anyway, so at the end this fact would be also a reference.
Uniparental markers may be tricky, I agree, but I was considering that most of them were males, by far. If I understand what you meant by looking at both sexes...

torzio
07-05-20, 19:35
Yeah, perhaps influences of groups such Celts and Rhaetians could help to explain it, but these are folks a bit older. Indeed, the NE is rich in R-L2 and relatively rich in G-L497. Plus, first Italics possibly had some resemblance to the formers, I believe, and if the latter were close to Etruscans also in Autosomal, then they would not have been that different from Italics either. That Roman soldier FN2 from 300 AD found in Bavaria was likely too late for a Rhaetian proper, but he could have been sort of a "Rhaetian relic", je je, since he was somewhat shifted to modern Iberians - just as Etruscans - and S. French IIRC, at the same time he carried a notorious "Alpine" Y marker. Plus, mountains are more prone to genetic continuity. So who knows!
"The Raeti tribes quickly became loyal subjects of the empire and contributed disproportionate numbers of recruits to the imperial Roman army's auxiliary corps."
(...)
"Further support for the hypothesis that the northern Raeti tribes converted to Celtic speech before the Roman imperial era is provided by the distribution of 'Raetian' inscriptions. These have been found almost exclusively in northeastern Italy: South Tyrol, Trentino, and the Veneto region. None have been found in Switzerland, the other core Raeti region. The Raetic inscriptions indicate that 'Raetian' survived as late as the 3rd century AD, suggesting that the Raeti tribes in this region at least may not have converted to Celtic speech." (Wikipedia)
Anyway, all of them must have been somewhat rich in EEF ancestry as well, even Celts (as per the shif in Iron Age UK). It's possibly someway involved in this "shared ancestry", naturally, while modern Italian cluster must also include other components, in turn not that strong in North Italy.
Finally, later movements probably helped to shape modern NE Italians, and we could include Germanics', sure, as well as movements from the South.
As for Germanics specifically, I said most part of that % in my mom could not be assigned to them, however, I agree that some of it could be. Not sure how much, but Y-DNA may provide clues in this regard. R-U106 must account for ~5% around Treviso province, and I1 frequency would be three times bigger in there. So perhaps something close to 20% around Treviso province specifically (where most of her ancestors were born). Now, if we think that the impact over Autosomal might have been substantially lower than that...
Thanks for the Eupedia's map on F&G. Notice that 42.2% is actually her NW, not her F&G, which is 34.8%. My father, on the other hand, scores 34.6 in the former and 24.5 in the latter, plus 1.6 East Euro, which is curious (my mom doesn't score any), also because he's basically 75% Venetian in ancestry, not fully. The other 25% come from far East Lombardy.

East Lombardy ie, major towns of Cremona, Brescia, Bergamo, Sondrio and up to the swiss order was under Venetian hands from 1430 to basically 1820, men served in venetian forces, went to venetian overseas colonies etc ................., when the Austrians took over after the congress of Vienna, they realigned the regional borders of Lombardy, Veneto and Friuli ( from 1820 to 1860/1870 ), not because of language or ethnicity , but to balance the land geographically............so the 25% of your fathers east-lombardy could also be venetian ...............all depends what what the programmers do

torzio
07-05-20, 19:42
for myself and family, very bad ethnic splits come from companies, Living DNA, Ancestry ....these are the worst ..............Ancestry actually gives me 43% french and 34% italian

myheritage gives me 70% Italian, 15% Balkan and 15% Irish .............nothing else

23andme give me 45% italian and 27% F&G

have you ..Riverman...any myhertiage numbers for yourself?

Riverman
07-05-20, 20:31
Well noticed. If my parents had that "Germanic" influence, we should perhaps expect, say, some Scandinavian %.

Most Germans have at least some Scandi, but not all. Oftentimes those with the highest F&G the least, because in them F&G eats up all the other subdivisions below NW. Its some kind of random assignment even, like I have written in the thread start.


My guess is that they were not "that" different from us, in part "accidentally", given those later influences in Veneto from both South and North. For a similar reason I think my great proximity to the "Illyrian" at MTA is in part "accidental". I'm certainly not an Illyrian relic. :)

That's certainly right, but question is how much of this old Italic/Etruscan being already in the Italian component? Seems its not that little.


Still concerning Germans, it may be true they don't score amazingly high in F&G, however, they score more than 90% in NW Euro, right? Very few S. Euro anyway, so at the end this fact would be also a reference.

Germans can vary a lot, some are about 80 percent F&G, others much less, but they are both German genealogically. The main variation in Germans is between NWE : EE, NWE : SE is just the 2nd most important variable.


Uniparental markers may be tricky, I agree, but I was considering that most of them were males, by far. If I understand what you meant by looking at both sexes...

Its always good to look at both. We know some Romans had wives from the North, even former slaves with which they had not just illegitimate, but also legitimate children. I guess one of the main reasons for the North : South differentiation in Imperial Rome was, that the North got more Northern, the South more South Eastern (East Mediterranean and further) immigrants, workers and slaves. I think some of the division started early and it was not just about old Italic-Etruscan vs. Greek-Phoenician colonies.

But the main change came with the Goths, Lombards, German and French (general Frankish) influence. F&G is, after all, just "Frankish" in the wider sense.


have you ..Riverman...any myhertiage numbers for yourself?

I posted some homogeneous group of Western Austrians without recent admixture, which score pretty much the same as Southern Bavarians across the border, here:
https://www.eupedia.com/forum/threads/33825-Share-your-23andMe-Ancestry-Composition/page17?p=602958&viewfull=1#post602958

They have much less broadly European than I have, which is obviously the result of 23andme's sampling and my personal mixture, which is not representative for a specific rural locality, but more typical for many urban areas.

Regio X
07-05-20, 21:09
East Lombardy ie, major towns of Cremona, Brescia, Bergamo, Sondrio and up to the swiss order was under Venetian hands from 1430 to basically 1820, men served in venetian forces, went to venetian overseas colonies etc ................., when the Austrians took over after the congress of Vienna, they realigned the regional borders of Lombardy, Veneto and Friuli ( from 1820 to 1860/1870 ), not because of language or ethnicity , but to balance the land geographically............so the 25% of your fathers east-lombardy could also be venetian ...............all depends what what the programmers doThanks, torzio. Interesting!
But what do you mean by programmers? I was talking on his actual ancestry. His paternal grandmother was born in province of Mantova, in a place close to Verona province. If her dialect was closer to veronese than to other Lombards', I don't know. Well, at least she could communicate with her husband (eastern) trevisano. :)


Most Germans have at least some Scandi, but not all. Oftentimes those with the highest F&G the least, because in them F&G eats up all the other subdivisions below NW. Its some kind of random assignment even, like I have written in the thread start.It makes sense. It's natural to think N. Germans tend to score more Scandi than, say, S. Germans. But I didn't know some Germans could get very few Scandi %. Interesting.


That's certainly right, but question is how much of this old Italic/Etruscan being already in the Italian component? Seems its not that little.Yes, some of it must be included in Italian component, sure. I don't know "how much".


Germans can vary a lot, some are about 80 percent F&G, others much less, but they are both German genealogically. The main variation in Germans is between NWE : EE, NWE : SE is just the 2nd most important variable.Indeed. East Germans in special, for instance, must get some relevant % of EE, but I won't suggest numbers because I actually don't remember of results.


Its always good to look at both. We know some Romans had wives from the North, even former slaves with which they had not just illegitimate, but also legitimate children. I guess one of the main reasons for the North : South differentiation in Imperial Rome was, that the North got more Northern, the South more South Eastern (East Mediterranean and further) immigrants, workers and slaves. I think some of the division started early and it was not just about old Italic-Etruscan vs. Greek-Phoenician colonies.

But the main change came with the Goths, Lombards, German and French (general Frankish) influence. F&G is, after all, just "Frankish" in the wider sense.

Riverman
07-05-20, 21:17
@Regio:
It makes sense. It's natural to think N. Germans tend to score more Scandi than, say, S. Germans. But I didn't know some Germans could get very few Scandi %. Interesting.

I saw a result on 23andme with probably one of the highest F&G of all and he scored zero = 0 Scandinavian. Among the Austrian-Bavarian sample too, one scored 0, like quoted in the thread start: 0-1,5 % Scandinavian. So a lot of German have 0 Scandinavian, especially the more Western ones and in a lot of them its pure chance, random assignment whether they got 0 or more Scandinavian. As a rule just subtracted from F&G and BNWE. Whether they are more Northern or not. I guess that's because Scandinavian being not just more Northern, but also somewhat more North Eastern shifted. The Eastern samples score more consistently Scandinavian.

That Italians don't have it probably might be attributed to the fact that they also have little to no of the North Eastern (hunter gatherer, Uralic?) ancestry, but exclusively the more Continental Germanic.

torzio
07-05-20, 21:37
Thanks, torzio. Interesting!
But what do you mean by programmers? I was talking on his actual ancestry. His paternal grandmother was born in province of Mantova, in a place close to Verona province. If her dialect was closer to veronese than to other Lombards', I don't know. Well, at least she could communicate with her husband (eastern) trevisano. :)

It makes sense. It's natural to think N. Germans tend to score more Scandi than, say, S. Germans. But I didn't know some Germans could get very few Scandi %. Interesting.

Yes, some of it must be included in Italian component, sure. I don't know "how much".

Indeed. East Germans in special, for instance, must get some relevant % of EE, but I won't suggest numbers because I actually don't remember of results.


many programmers use today's borders as a ethnicity marker and do not go back in time to check the history , even recent history...............I,my opinion, never go back beyond medieval times for ethnicity as I think before this period, be it with the Roman empire and barbarian invasions, ethnicity in europe was chaotic

Regio X
07-05-20, 21:38
@Regio:

I saw a result on 23andme with probably one of the highest F&G of all and he scored zero = 0 Scandinavian. Among the Austrian-Bavarian sample too, one scored 0, like quoted in the thread start: 0-1,5 % Scandinavian. So a lot of German have 0 Scandinavian, especially the more Western ones and in a lot of them its pure chance, random assignment whether they got 0 or more Scandinavian. As a rule just subtracted from F&G and BNWE. Whether they are more Northern or not. I guess that's because Scandinavian being not just more Northern, but also somewhat more North Eastern shifted. The Eastern samples score more consistently Scandinavian.Ops. I didn't read it carefully, and I probably missed that detail.


That Italians don't have it probably might be attributed to the fact that they also have little to no of the North Eastern (hunter gatherer, Uralic?) ancestry, but exclusively the more Continental Germanic.Based on the informations you provided, it looks to be the case, yes, but that F&G in N. Italians must be indeed mostly of non-Germanic origin anyway.

torzio
07-05-20, 21:46
Thanks, torzio. Interesting!
But what do you mean by programmers? I was talking on his actual ancestry. His paternal grandmother was born in province of Mantova, in a place close to Verona province. If her dialect was closer to veronese than to other Lombards', I don't know. Well, at least she could communicate with her husband (eastern) trevisano. :)

It makes sense. It's natural to think N. Germans tend to score more Scandi than, say, S. Germans. But I didn't know some Germans could get very few Scandi %. Interesting.

Yes, some of it must be included in Italian component, sure. I don't know "how much".

Indeed. East Germans in special, for instance, must get some relevant % of EE, but I won't suggest numbers because I actually don't remember of results.

Mantua is interesting .................there is an interesting book on mantua and its people and ethnic history, called......a renaissance tapestry, by Kate Simon..........basically there was a lot of mixing with the people of Nassau https://en.wikipedia.org/wiki/Nassau,_Rhineland-Palatinate ................normally I would not care as this marriage mixing would count for nothing, but when a populace is small and you get Nassau mercenaries staying , then there can be a greater percent of other ethnicity

Verona is a province of the region of Veneto ............I do not know what you mean by Lombard

Regio X
07-05-20, 22:02
Mantua is interesting .................there is an interesting book on mantua and its people and ethnic history, called......a renaissance tapestry, by Kate Simon..........basically there was a lot of mixing with the people of Nassau https://en.wikipedia.org/wiki/Nassau,_Rhineland-Palatinate ................normally I would not care as this marriage mixing would count for nothing, but when a populace is small and you get Nassau mercenaries staying , then there can be a greater percent of other ethnicity

Verona is a province of the region of Veneto ............I do not know what you mean by LombardI know Verona is in Veneto. :)
His grandmother was born up to 5 miles from Verona province. I questioned if her dialect could be closer to the dialect of Verona than to other Lombard dialects (since Mantova province is in Lombardy), already assuming that language may be a reference of "(sub)ethnicity". For example, the dialect spoken in Caneva-PN is the trevisani, certainly different from, say, the udinesi, never mind the fact the comune is in the region of Friuli-Venezia Giulia.

Riverman
07-05-20, 22:32
was a lot of mixing with the people of Nassau https://en.wikipedia.org/wiki/Nassau,_Rhineland-Palatinate

Not that I think they alone did it, but just to mention it: The people of Nassau and the Rhineland-Palatinate in general, if not having recent admixture from elsewhere, should be among the Germans with the highest F&G percentages imho. At least if the composition algorithm works correctly.

torzio
07-05-20, 22:32
I know Verona is in Veneto. :)
His grandmother was born up to 5 miles from Verona province. I questioned if her dialect could be closer to the dialect of Verona than to other Lombard dialects (since Mantova province is in Lombardy), already assuming that language may be a reference of "(sub)ethnicity". For example, the dialect spoken in Caneva-PN is the trevisani, certainly different from, say, the udinesi, never mind the fact the comune is in the region of Friuli-Venezia Giulia.


Caneva is in Pordenone area ...........Pordenone was in Veneto lands until the Austrians shifted the border of Veneto and Friuli to the Livenza river. you will find a very low % of people in Pordenone speak friulian , the most would be in Azzano Decimo ..............most people in Pordenone speak Veneto .
I have a few relatives in Pordenone province , one in Prata and other in Porcia........they only speak Veneto

The only province where Friulian dominates is Udine province

Even Trieste , which was never ever under Venice , it populace speak more higher percent Veneto than Friulian

Regio X
07-05-20, 23:11
Caneva is in Pordenone area ...........Pordenone was in Veneto lands until the Austrians shifted the border of Veneto and Friuli to the Livenza river. you will find a very low % of people in Pordenone speak friulian , the most would be in Azzano Decimo ..............most people in Pordenone speak Veneto .
I have a few relatives in Pordenone province , one in Prata and other in Porcia........they only speak VenetoYes, that was the point. I referred to my father's grandmother place in this sense, given the proximity to Verona province.

Duarte
08-05-20, 00:32
That Roman soldier FN2 from 300 AD found in Bavaria was likely too late for a Rhaetian proper, but he could have been sort of a "Rhaetian relic", je je, since he was somewhat shifted to modern Iberians

Hi dear fellow Regio. You are absolutely right. The FN_2 is an Iberian like. Cheers ;)

https://i.imgur.com/WeT4BWi.jpg

https://i.imgur.com/1Pp1364.jpg
https://i.imgur.com/HMvw8tW.jpg

https://i.imgur.com/bNPehU9.jpg
https://i.imgur.com/NF0zcZT.jpg

Sources:
https://www.pnas.org/content/115/13/3494

https://www.pnas.org/content/pnas/suppl/2018/03/06/1719880115.DCSupplemental/pnas.1719880115.sapp.pdf

torzio
08-05-20, 00:47
Hi dear fellow Regio. You are absolutely right. The FN_2 is an Iberian like. Cheers ;)

https://i.imgur.com/WeT4BWi.jpg

https://i.imgur.com/1Pp1364.jpg
https://i.imgur.com/HMvw8tW.jpg

https://i.imgur.com/bNPehU9.jpg
https://i.imgur.com/NF0zcZT.jpg

Sources:
https://www.pnas.org/content/115/13/3494

https://www.pnas.org/content/pnas/suppl/2018/03/06/1719880115.DCSupplemental/pnas.1719880115.sapp.pdf


Roman soldier sampled from Munich (FN_2). We find some significant differences in regional ancestry comparing the Altheim individuals to FN2. For example, FN2 exhibits substantially more West Asian-like ancestry than ALH_1 and ALH_10 (Fig. S48), also consistent with Fig. S44-S46, and substantially more Middle Eastern-like and Southern European-like ancestry.

FN_2, and VIM_2 are lactase persistent.

Duarte
08-05-20, 01:24
Roman soldier sampled from Munich (FN_2). We find some significant differences in regional ancestry comparing the Altheim individuals to FN2. For example, FN2 exhibits substantially more West Asian-like ancestry than ALH_1 and ALH_10 (Fig. S48), also consistent with Fig. S44-S46, and substantially more Middle Eastern-like and Southern European-like ancestry.

FN_2, and VIM_2 are lactase persistent.


Hi @torzio.
Most likely he was a showy Roman soldier in his beautiful legionary’s uniform. The women with deformed skull must have been enchanted by him, but I very much doubt that he was enchanted by them. :grin:
Cheers :good_job::smile:

Regio X
08-05-20, 04:04
Hi dear fellow Regio. You are absolutely right. The FN_2 is an Iberian like. Cheers ;)

https://i.imgur.com/WeT4BWi.jpg

https://i.imgur.com/1Pp1364.jpg
https://i.imgur.com/HMvw8tW.jpg

https://i.imgur.com/bNPehU9.jpg
https://i.imgur.com/NF0zcZT.jpg

Sources:
https://www.pnas.org/content/115/13/3494

https://www.pnas.org/content/pnas/suppl/2018/03/06/1719880115.DCSupplemental/pnas.1719880115.sapp.pdfYep, it's explicitly in the paper:
"This is in contrast to the Roman soldier dating to around 300 AD sampled from the same region, for which its largest ancestry component was IBS, with greatest genetic similarity to modern Spanish and southern French individuals (SI Appendix, Fig. S31 (http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1719880115/-/DCSupplemental)). Based on an analysis of patterns of haplotype sharing, the Roman soldier (FN_2: 11.08×) was found to have substantially more southern European, West Asian, and Middle Eastern ancestry than two normal-skulled Early Medieval Bavarians with high genomic coverage"

Notice they say modern Spanish and southern French. Not sure why they assumed a necessary SW ancestry in another part of the text then, considering these almost 2000 years.
"The analysis of his genome identifies him to be of southwest European origin."

How can they be so sure? It'd be like thinking Etruscans came from SW Euro because of genetic resemblance to modern Iberians. (It seems Etruscans were actually between modern Iberians and N. Italians. I'd have to re-check it. Still...)

That said, assuming the same approach was used to estimate their genetic resemblance to modern pops, I cogitated that FN2 was Rhaetian-like because of this same Etruscan shift toward modern Iberians, at the same time Rhaetian and Etruscan languages (non-IEs) were also related to one another, not to mention the area where the sample was found. Coincidence? Maybe it is, maybe it isn't. Plus, the soldier carried G-L42, the "Alpine" Y marker I mentioned (interestingly, the only basal G-Z40854 - his subclade - thus far is from Udine). Possibly another coincidence. :) I know the Etruscans samples are much older, but if the "late" Rhaetians - especially in Italy - were really not that different, I wouldn't be so surprised, after all, some genetic continuity may be expected in mountainous areas, and at the end "Italics" in NE Italy could be also similar to some Rhaetians (so an occasional mix would not necessarily result in a "that different" third group).
Now, I see these facts as clues. Truth is that we need ancient Rhaetian DNA to confirm or negate this possible feature of Rhaetian people (or at least part of them) as being close to modern Spanish and S. French in Autosomal.