German GEDmatch results

[mariusz99]

Stearsolina, do you have in plans to update Polish/ Ukrainian averages or do you think they are good and need no update?

[Leto]

Stearsolina, do you have in plans to update Polish/ Ukrainian averages or do you think they are good and need no update?

vbnetkhio did that last year. Poland is more homogeneous than Germany. Also a lot of resettled people from East to West.

[Mejgusu]

Its interesting that many Germans are genetically Scandinavian+Iberian influence. I expected Scandinavian+Eastern influence in Eastern Germans but not this in Western Germans.

[vbnetkhio]

It's kind of odd. Carinthia and Styria have visible genetic profile that is Slovenian like. Burgenland is surprisingly different, not so Slavic/south Slavic and more "German shifted" but with notable Slavic also.
Tyrol looks Swiss like and between SW Germans and Alpine Italians. That's all fine.

Lower and Upper Austria/Salzburg are mess. Some are SW German like, some are Swiss like, some are Carinthian/Slavic shifted. And these people have deep ancestry in these areas.
We are trying to get few more samples from these regions to estimate what's best to do.

but is there any kind of pattern? maybe a more complicated one, like those closer to Styria and Czechia are both more Slavic, and those in the middle less? if not, just make a geographical Upper and Lower Austrian average.

[Jana]

Okay. My problem is that when there are too many (similar) averages it will cause overfitting or unrealistic models. For example the Brandenburg outlier already popped up in some Slavo-Germanic results. I believe half Volga half Russian or something like that.

Please keep the original German names, they are more colorful IMO "Pflalz" is better than "Palatinate" for ex.

That Brandenburg result is heavy outlier, but fully native there and has only German names in last 200 years if not more! That why it important to keep it, likely one smaller entire area of that state plots like that (to get such results with such continuity there must mean other people/Germans around were genetically like that as well).

Btw Davidski uses that sample as example of Sorb, and that person isn't Sorb at all. It has only German names in family tree. Native Brandenburg kits are nightmare to find btw, rothaer searches them for years and nothing.
As for names, agree German names sound better. We wanted them to be more understandable for others, but if others don't mind, can't see why we couldn't use them.

Smaller averages are just when we aren't sure or expect genetic differences in rest of same state/region. For example there is Saxony Ore Mountains average because it was rather different from academic Leipzig samples we had, but in new average it will be gone, other verified Saxons aren't too different from them. Same for BW state, etc.

[vbnetkhio]

Stearsolina, do you have in plans to update Polish/ Ukrainian averages or do you think they are good and need no update?

this is the final version of mine and Peterski's averages:

Poland_Lemko,27.74,37.63,14.89,7.34,9.48,0.24,0.42,0.58,0.56,0.60,0.52,0.00,0.00
Polish_Greater_Poland,31.45,43.92,10.26,5.30,5.52,0.80,0.87,0.09,0.23,0.65,0.48,0.37,0.06
Polish_Kielce,29.3,45.17,10.56,5.45,5.36,0.76,1.36,0.15,0.37,0.81,0.4,0.28,0.03
Polish_Kujawy,31.83,45.41,9.95,4.66,3.33,1.47,1.02,0.20,0.79,0.93,0.00,0.41,0.00
Polish_Lesser_Poland,31.00,43.23,11.15,5.26,5.63,0.55,0.47,0.06,0.80,0.85,0.57,0.29,0.16
Polish_Mazovia,28.33,46.20,10.70,5.60,5.04,0.54,1.20,0.15,0.69,0.66,0.50,0.18,0.20
Polish_North_Wielkopolska,31.98,43.23,10.06,5.79,5.57,0.87,0.73,0.17,0.19,1.00,0.03,0.35,0.03
Polish_Podlaskie,29.87,46.68,7.30,4.42,6.83,1.65,1.33,0.00,1.06,0.35,0.52,0.00,0.00
Polish_Silesia,29.25,43.03,12.16,6.16,5.57,0.39,1.25,0.29,0.64,0.70,0.42,0.16,0.00
Polish_Southeast,28.71,42.67,11.10,6.04,7.21,1.18,0.44,0.11,1.70,0.23,0.12,0.35,0.13
Polish_Staropolska,29.08,46.18,10.22,4.40,6.59,0.77,0.73,0.00,0.45,0.75,0.37,0.46,0.00
Polish_WarmiaMasuria,32.51,44.26,10.88,4.19,3.27,0.95,1.21,0.16,1.06,0.51,0.33,0.29,0.39

this is not exactly the same as those on Vahaduo now, I don't know what happened there.

[Jana]

but is there any kind of pattern? maybe a more complicated one, like those closer to Styria and Czechia are both more Slavic, and those in the middle less? if not, just make a geographical Upper and Lower Austrian average.

ohh geographic pattern isn't clear yet, still need more kits. And some samples look "too northern" for their location (upper Austrian plotting with Rhinelanders for eg. being totally native there)
than you have Noricum sis in law from lower Austria that's supposed to be more Slavic than upper Austrians and she has Swiss/Tyrol/south shifted Bavarian shift while one Upper Austrian has heavy Slav shift!

a mess.

[Leto]

What is the origin of Davidski's Austrian average? And how many samples were used back then?

[Jana]

What is the origin of Davidski's Austrian average? And how many samples were used back then?

No idea, probably people from Vienna with various backgrounds. It doesn't look very accurate based on results we have so far, too much eastern.
Also working to replace Slovenian and Czech academic averages (Slovenian current one is pretty fine though, Czech less so)

[rothaer]

I don't think g25 has this problem.. k36 does, for example it has a fennoscandian component based on modern Finns, so it can model Finns properly, but it still can't deal with Sami (because of a lack of WHG and EHG components)

G25 has the problem that Northern Easter Germans do get loads of Swedish, which is due to a little Finnish or Finnish-like input in Swedes, that is confused with a Germanic-Slavic mixture. I have two G25 coords where that excessively happens. While this confusion does not at all happen in K36.