Eastern European alternative G25 genetic model

[AndreAndre]

Since Russki ( a user on this forum) has made his Estonian-Albanian model, I started wondering how come Western Europeans score so much albanian on that model
The solution is very easy to understand - because Albanians are a more western shifted population than are Estonians.
Now if we were to model East Europeans as a mixture of Lithuanians (a relatively western group) and Crimean Tatars (a relatively eastern group), we would get rather different results.
I dont claim that my model is superior, it is just an alternative to what Russki offered.
Target: AndreiK_scaled
Distance: 0.0267% / 0.02670564
84.0 Lithuanian_PA
16.0 Tatar_Crimean_steppe

Target: Polish
Distance: 0.0257% / 0.02571865
96.1 Lithuanian_PA
3.9 Tatar_Crimean_steppe

Target: Ukrainian_Lviv
Distance: 0.0365% / 0.03650269
92.5 Lithuanian_PA
7.5 Tatar_Crimean_steppe

Target: Russian_Yaroslavl
Distance: 0.0202% / 0.02017335
89.9 Lithuanian_PA
10.1 Tatar_Crimean_steppe

Target: Russian_Orel
Distance: 0.0164% / 0.01637027
91.9 Lithuanian_PA
8.1 Tatar_Crimean_steppe

Target: Belarusian
Distance: 0.0126% / 0.01258312
98.0 Lithuanian_PA
2.0 Tatar_Crimean_steppe

[Russki]

We can model Estonians as some Lithuanian-Tatar mix

We can model Belarusians as relatively pure Lithuanians

But for what?

https://sun9-37.userapi.com/impf/u-cuudl9dED54dI7N9dAYr2vP65mBlfKLGlHaw/x1Z4PgLa9jE.jpg?size=885x379&quality=95&sign=6f4dd8dcd05d6aed3b545f1f42fe88f4&type=album




Albanians are more Western shifted than Estonians and that model was designed only for Eastern Europeans

For Western Europeans it's not very accurate

[AndreAndre]

Now, adding Albanians to this two way model to make it a 3 way model produces more accurate results.
Here are the results with Albanians added (this shows just how Southwestern and Southeastern shift these populations have)
Target: Belarusian
Distance: 0.0109% / 0.01090318
94.2 Lithuanian_PA
5.1 Albanian
0.7 Tatar_Crimean_steppe

We see that belarusians are rather northern, with minor southern and even more minor eastern shift
Target: Russian_Orel
Distance: 0.0114% / 0.01142802
84.8 Lithuanian_PA
9.5 Albanian
5.7 Tatar_Crimean_steppe

Russians of Oryol are less northern

Target: Russian_Yaroslavl
Distance: 0.0202% / 0.02017335
89.9 Lithuanian_PA
10.1 Tatar_Crimean_steppe

Russians of Yaroslavl are not very northern but they are very eastern, and lack southern shift

Target: Ukrainian_Lviv
Distance: 0.0130% / 0.01295761
71.8 Lithuanian_PA
27.7 Albanian
0.5 Tatar_Crimean_steppe

Ukrainians of lviv are very southern but not very eastern
Target: Polish
Distance: 0.0111% / 0.01105697
81.7 Lithuanian_PA
18.3 Albanian

Polacks are less southern that ukrainians from lviv
Target: AndreiK_scaled
Distance: 0.0267% / 0.02670564
84.0 Lithuanian_PA
16.0 Tatar_Crimean_steppe

I am less southern than polacks are, and lack southwestern/albanian pull

[AndreAndre]

Albanians are more Western shifted than Estonians and that model was designed only for Eastern Europeans

For Western Europeans it's not very accurate

For eastern europeans it is also not very accurate. The more "western" eastern euros will get larger Albanian scores and lesser Estonian scores, and tighter fits. Whereas Russians (who have mostly southeastern pull relative to estonians) will get higher estonian (due to estonians being an eastern population). Your model makes it seem like Russians are northerners relative to Poles, when in reality that isnt the case.

[Russki]

For eastern europeans it is also not very accurate. The more "western" eastern euros will get larger Albanian scores and lesser Estonian scores, and tighter fits.


I modeled Ukrainian_Lviv and Polish with distances 0.014 and 0.012, whereas you modeled them with distances 0.037 and 0.026.

My distances are over twice shorter, which makes my model better than yours.




Your model makes it seem like Russians are northerners relative to Poles, when in reality that isnt the case.

I don't use Lithuanians as a reference point because their admixture is more Northern than Estonians — they are both more WHG and less Anatolian Farmer than them.

https://sun9-46.userapi.com/impf/e-lzro_xhvZdk6UaISouuE0Ri5rsMQhSFTyGlA/CawbMuXQz28.jpg?size=1080x446&quality=95&sign=4f3d7a111d7e4479e6bceec2323c6168&type=album



It doesn't make any sense if you realize that Estonia is farther North than Lithuania, and on a G25 PCA they are more Northern too.

It starts to make sense only if you release ANE, which is a very Northern-shifted element, and is considerably bigger in Estonians, Russians (all) and Finns than in other populations.


No matter how you spin Global25 models, on many of which Lithuanians are more Northern-admixed than Estonians and Belarusians are more Northern-admixed than Russians, the result on Global25 PCA will be permanently the same.

Poles on G25 cluster with Southernmost Russians, on Belgorod-Voronezh line.

https://sun9-12.userapi.com/impf/UklmreOkYq7P_v_C_w8U_OgEBATYNWteolgoZQ/zLo1L4VOxK0.jpg?size=1537x685&quality=95&sign=9116b1d26d962af47507d50cda3a6662&type=album

[Russki]

Whereas Russians (who have mostly southeastern pull relative to estonians) will get higher estonian (due to estonians being an eastern population).



Up to ~tver Russians are more western-shifted than Estonians.

You can check it with Swedes or any other NW Euro group.

https://sun9-83.userapi.com/impf/-pxozsAb8PPTwhVJw--kmC0s52dnnZGCTzYhMQ/Rw-igq0WP2M.jpg?size=1080x585&quality=95&sign=956f8ce864c799a3f318894bec7d8850&type=album



The best for it are Nordic ethnicities due to minimal Southern pull.

[AndreAndre]

Up to ~tver Russians are more western-shifted than Estonians.

You can check it with Swedes or any other NW Euro group.

https://sun9-83.userapi.com/impf/-pxozsAb8PPTwhVJw--kmC0s52dnnZGCTzYhMQ/Rw-igq0WP2M.jpg?size=1080x585&quality=95&sign=956f8ce864c799a3f318894bec7d8850&type=album



The best for it are Nordic ethnicities due to minimal Southern pull.

You compared Russians from Yaroslavl to Poles to Ukrainians from Lviv in an earlier post. Yaroslav Russians are east of Estonians
Distance to: Mansi
0.32080541 Russian_Yaroslavl
0.33326744 Estonian

Distance to: Tatar_Crimean_steppe
0.18920357 Russian_Yaroslavl
0.20632956 Estonian

Russian Yaroslavl modeled as Estonian + Albanian produces a worse fit that Estonian + Crimean Tatar
Target: Russian_Yaroslavl
Distance: 0.0197% / 0.01973555
91.2 Estonian
8.8 Tatar_Crimean_steppe
Target: Russian_Yaroslavl
Distance: 0.0219% / 0.02193206
89.5 Estonian
10.5 Albanian

So my model is indeed better for this situation.