Who is more Mongoloid on average: Hungarians or Estonians?

[Zanzibar]

Just curious here. Who has more East Eurasian ancestry on average? From what I know both groups have negligible amounts of East Eurasian admixture.

[Halvard]

I'd say Estonians since they have more ANE admixture.

[Alenka]

Estonians are more Mongoloid.

Eurogenes K13: (https://docs.google.com/spreadsheets/d/1dCZldTIfd-EPjDlpQiFNcHwOtZus9Qdll3pB48zdQG0/edit#gid=0)
Estonian
Siberian: 3.51%
Amerindian: 1.35%
East Asian: 0.05%
Oceanian: 0.61%

Hungarian
Siberian: 1.11%
Amerindian: 0.97%
East Asian: 0.21%
Oceanian: 0.14%

[Zanzibar]

Estonians are more Mongoloid.

Eurogenes K13: (https://docs.google.com/spreadsheets/d/1dCZldTIfd-EPjDlpQiFNcHwOtZus9Qdll3pB48zdQG0/edit#gid=0)
Estonian
Siberian: 3.51%
Amerindian: 1.35%
East Asian: 0.05%
Oceanian: 0.61%

Hungarian
Siberian: 1.11%
Amerindian: 0.97%
East Asian: 0.21%
Oceanian: 0.14%

Interesting. Could this be because Estonians mixed less with their neighbors than Hungarians thus preserving their Mongoloid ancestry better?

I wonder if Csango and Szekely would score more Mongoloid than the average Hungarian.

[Ion Basescul]

Interesting. Could this be because Estonians mixed less with their neighbors than Hungarians thus preserving their Mongoloid ancestry better?

I wonder if Csango and Szekely would score more Mongoloid than the average Hungarian.

Csangos can score solid amounts. They are the best preserved Hungarian group, judging by the East Asian-like ancestry. The irony is that most identify as Romanians of Catholic rite nowadays.

<google-sheets-html-origin style="color: rgb(0, 0, 0); font-size: medium;">
<tbody>
East_Asian
Siberian
Amerindian
Total


0
2.96
0.42
3.38


1.40
3.06
1.43
5.89


1.52
1.14
1.49
4.15


1.44
3.26
0
4.70


0.98
2.84
0.20
4.02


0
1.77
0.35
2.12


0
1.60
1.41
3.01


0
0.91
0
0.91


4.67
0.40
2.40
7.47


3.02
3.06
0
6.08


2.93
1.70
0.99
5.62


0.14
2.89
0.79
3.82


1.71
3.99
1.59
7.29


1.39
2.63
0.88
4.90


0
3.35
0
3.35


0
3.31
0
3.31


0.45
1.88
1.81
4.14


0.49
2.66
1.34
4.49


1.80
3.81
1.32
6.93


1.51
3.17
0.16
4.84


0
3.09
0.00
3.09


1.08
2.53
0.59
4.20


1.12
2.22
0.87
4.21


0
3.02
1.67
4.69


1.53
3.10
0.56
5.19


1.80
2.02
0.81
4.63


1.37
2.78
2.38
6.53

</tbody>
</google-sheets-html-origin>

[Zanzibar]

Csangos can score solid amounts. They are the best preserved Hungarian group, judging by the East Asian-like ancestry. The irony is that most identify as Romanians of Catholic rite nowadays.

<google-sheets-html-origin style="color: rgb(0, 0, 0); font-size: medium;">
<tbody>
East_Asian
Siberian
Amerindian
Total


0
2.96
0.42
3.38


1.40
3.06
1.43
5.89


1.52
1.14
1.49
4.15


1.44
3.26
0
4.70


0.98
2.84
0.20
4.02


0
1.77
0.35
2.12


0
1.60
1.41
3.01


0
0.91
0
0.91


4.67
0.40
2.40
7.47


3.02
3.06
0
6.08


2.93
1.70
0.99
5.62


0.14
2.89
0.79
3.82


1.71
3.99
1.59
7.29


1.39
2.63
0.88
4.90


0
3.35
0
3.35


0
3.31
0
3.31


0.45
1.88
1.81
4.14


0.49
2.66
1.34
4.49


1.80
3.81
1.32
6.93


1.51
3.17
0.16
4.84


0
3.09
0.00
3.09


1.08
2.53
0.59
4.20


1.12
2.22
0.87
4.21


0
3.02
1.67
4.69


1.53
3.10
0.56
5.19


1.80
2.02
0.81
4.63


1.37
2.78
2.38
6.53

</tbody>
</google-sheets-html-origin>

How about Szekely? Can a Csango score 10% East Asian or that's too high?

5-7% Mongoloid is pretty high for a Euro, that's close to Finnish-tier level of Mong.

[Kökény]

How about Szekely? Can a Csango score 10% East Asian or that's too high?

5-7% Mongoloid is pretty high for a Euro, that's Finnish-tier level of Mong.
My father scores higher than the Estonian average (he's between 5-6%). Székely and Csángó mong scores are quite close, although there are Csángó outliers who reach up to 7%.

I don't think they can reach 10% though. That's Balkan Turk level.

[Zanzibar]

My father scores higher than the Estonian average (he's between 5-6%). Székely and Csángó mong scores are quite close, although there are Csángó outliers who reach up to 7%.

I don't think they can reach 10% though. That's Balkan Turk level.

Hmm fascinating. How much do you score btw?

It would be interesting if there are also G25 Csango and Szekely samples to compare. From What I remember the highest East Asian I see for an Estonian sample from G25 is 5.6% (all the others were around 1-2%) maybe he/she has recent Finnish origin.

Do you think its direct ancestry from the Magyar conquerors or there were some later gene flows from Turkic invasions as well?

I see. Actually the several Finns who have around 10% I East Eurasian believe.

[Kökény]

Hmm fascinating. How much do you score btw?
Close to 5% (4.73%).

Do you think its direct ancestry from the Magyar conquerors or there were some later gene flows from Turkic invasions as well?
Could be both, I'm not sure. In the case of Székelys there are theories about Turkic origin rather than Magyar. Some people think it's BS. Personally, I don't entirely reject those theories.

[Lemminkäinen]

Interesting. Could this be because Estonians mixed less with their neighbors than Hungarians thus preserving their Mongoloid ancestry better?

I wonder if Csango and Szekely would score more Mongoloid than the average Hungarian.

Conversely, Estonians have same ancient Saami admixture than we see in Finland, due to the Finnish migration from the Late Iron Age. Early Iron Age Estonians had near zero Siberian.

[Lemminkäinen]

Seems like the Finns, descendants of ancient Estonians and Scandinavians, got mixed with Saamis and brought a minor Siberian to Estonia and Sweden. We have an alternative theory claiming that Baltic Finns came from the Volga Ural region carrying high amount of Siberian, but recently published ancient dna doesn't support it. Bad, Finnish researchers are very shy and don't want to publish dna finds from the Iron Age Eura.

The answer is, Iron Age Estonians had probably less Siberian than Hungarian speakers at the same time. Today Estonians have more.

[Petalpusher]

at K3:

Estonian 5.40 0.00 94.60

Hungarian 3.57 0.08 96.36

Although it also includes Oceanian and some part of Amerindian/ANE. Anything in the eastern half of Eurasia, or not LBK like.

[Zanzibar]

Close to 5% (4.73%).

Could be both, I'm not sure. In the case of Székelys there are theories about Turkic origin rather than Magyar. Some people think it's BS. Personally, I don't entirely reject those theories.

I see. What's the highest Mongoloid a Hungarian who is neither a Csango or Szekely can score?

Are there ancient Magyar samples that has been found and how much Mong and Caucasoid they have?

Here are the G25 Estonian individuals, the highest East Eurasian score here is 5.6%. Krasnoyarsk_BA represents the Siberian ancestry of Uralics. I decided to use Danish instead of Swedish because the latter seems to have significant Baltic admixture while Danish seems to be purer Germanic. This is to see if Estonians have any non-Baltic admixture and also to improve the distance fits.

https://i.imgur.com/tNdoiSu.png

Now compared this to the Hungarian samples: please correct me if I am modelling it wrong. This is what I presume is the genetic component of Hungarians. The highest East Eurasian score here is 2.6%. I used Tuvinian (Turkic) instead of Krasnoyarsk_BA because it seems to give a bit higher East Eurasian score.

https://i.imgur.com/NDBpQcJ.png

Here is another model I attempted: https://i.imgur.com/vpVbI5g.png

What do you think?

[Zanzibar]

Seems like the Finns, descendants of ancient Estonians and Scandinavians, got mixed with Saamis and brought a minor Siberian to Estonia and Sweden. We have an alternative theory claiming that Baltic Finns came from the Volga Ural region carrying high amount of Siberian, but recently published ancient dna doesn't support it. Bad, Finnish researchers are very shy and don't want to publish dna finds from the Iron Age Eura.

The answer is, Iron Age Estonians had probably less Siberian than Hungarian speakers at the same time. Today Estonians have more.

Interesting. How much do Estonians have in terms of East Eurasian ancestry? Do you know why Estonians have lower Siberian than Finns on average? Swedes also have minor East Eurasian admixture?

[Blondie]

Hungarians are such central european population like everyone else in the region, except székelys and csángós they are more eastern shifted, but these groups are only 5% of hungarians.

[Zanzibar]

Hungarians are such central european population like everyone else in the region, except székelys and csángós they are more eastern shifted, but these groups are only 5% of hungarians.

So I guess Estonians overall have more Mongoloid admixture than Hungarians?

Why is there very little to almost none East Eurasian ancestry in modern day Hungarians?

[Blondie]

So I guess Estonians overall have more Mongoloid admixture than Hungarians?

Yes definitely.


Why is there very little to almost none East Eurasian ancestry in modern day Hungarians?

Because nomad magyars were only 20000-50000 people and locals were 1 million, in the 9. century. And in the last 1000 years lot of things happened: tons of migration waves from other european countries, mongol genocide, ottoman genocide, many wars etc. If you go to Hungary, Slovakia, Czechia, Croatia, Slovenia etc you will see same faces everywere.

[Zanzibar]

Yes definitely.

Because nomad magyars were only 20000-50000 people and locals were 1 million, in the 9. century. And in the last 1000 years lot of things happened: tons of migration waves from other european countries, mongol genocide, ottoman genocide, many wars etc. If you go to Hungary, Slovakia, Czechia, Croatia, Slovenia etc you will see same faces everywere.

Make sense, Estonians are more isolated from other Euros and other internal migrations and movements than Hungarians, I guess.

What's the highest amount of Mongoloid you have seen for a non-Csango/non-Szekely Hungarian?

How much Mong do the original Magyars have?

[Hektor12]

Close to 5% (4.73%).Is this visible in your phenotype?

[Komintasavalta]

at K3:

Estonian 5.40 0.00 94.60

Hungarian 3.57 0.08 96.36

Although it also includes Oceanian and some part of Amerindian/ANE. Anything in the eastern half of Eurasia, or not LBK like.

I did a K=3 ADMIXTURE run that only included Eurasian populations. Now Estonians had 9% of the Siberian component and Hungarians had 6%.

Actually even Estonians probably have over 10% Mongoloid ancestry (or undifferentiated Eurasian ancestry that is comparable to Mongoloid ancestry), but this run included only modern populations, so part of the Mongoloid ancestry in steppe and EHG is included in the Caucasoid component.

https://i.ibb.co/9WmgDg5/circlize-eurasia-k3.jpg

The clustering is based on a combined matrix of admixture weights at different K values, so for example Cretans and Sardinians are in different clusters even though both of them have 100% of the Caucasoid component at K=3.

[Lemminkäinen]

Interesting. How much do Estonians have in terms of East Eurasian ancestry? Do you know why Estonians have lower Siberian than Finns on average? Swedes also have minor East Eurasian admixture?

As I wrote, the Finns have Iron Age Saami admixture. This is not a conspiracy theory, but supported by Finnish historians, old toponyms, arhaeology, linguist reseach and now even by ancient genome samples. I have not seen evidences about anything else. Those who connect Baltic Finnic languages to Siberian admixture are not aware of the difference between causation and correlation. Those who insist something else are conspiracy theorists or prefer politics.

The Finnish migration to Estonia is a widely known fact.

[Hektor12]

As I wrote, the Finns have Iron Age Saami admixture.Can you clarify whats the iron age Saami? As origins and components?

[Petalpusher]

I did a K=3 ADMIXTURE run that only included Eurasian populations. Now Estonians had 9% of the Siberian component and Hungarians had 6%.

Actually even Estonians probably have over 10% Mongoloid ancestry (or undifferentiated Eurasian ancestry that is comparable to Mongoloid ancestry), but this run included only modern populations, so part of the Mongoloid ancestry in steppe and EHG is included in the Caucasoid component.

https://i.ibb.co/9WmgDg5/circlize-eurasia-k3.jpg

The clustering is based on a combined matrix of admixture weights at different K values, so for example Cretans and Sardinians are in different clusters even though both of them have 100% of the Caucasoid component at K=3.

I would try that run with Oceanian and SSA, this is really what is missing in the Gedrosia K3. In south Indians groups, it's not solely East Asian stuff in there but at least some ancient ENA/Oceanian thing too. Also middle easterns have various real low levels of SSA but they likely have a real Eastern Eurasia element too

[Zoro]

I did a K=3 ADMIXTURE run that only included Eurasian populations. Now Estonians had 9% of the Siberian component and Hungarians had 6%.

Actually even Estonians probably have over 10% Mongoloid ancestry (or undifferentiated Eurasian ancestry that is comparable to Mongoloid ancestry), but this run included only modern populations, so part of the Mongoloid ancestry in steppe and EHG is included in the Caucasoid component.

https://i.ibb.co/9WmgDg5/circlize-eurasia-k3.jpg

The clustering is based on a combined matrix of admixture weights at different K values, so for example Cretans and Sardinians are in different clusters even though both of them have 100% of the Caucasoid component at K=3.

It’s good to see that at least you and Petalpusher understand that you can’t use a k9 or k11 or k13 calculator to figure out total East Eurasian. The only thing I would suggest is that you use 3 separate West Eurasian components; ENF, WHG, Taforalt and add them for combined West Eurasian

[Lemminkäinen]

Can you clarify whats the iron age Saami? As origins and components?

Yeah. The IA connection between Finns and Saamis have been known to have existed in Southern Finland before the Saamis' northern migration or how it is said in other words, before the Finns pushed them to north. We have tons of hard scientific evidences about it. So it is not a surprise that Levanluhta samples turned out to be heavily Siperian and very close modern Saamis, they even didn't have to be, but they were. All this science is not a new find, but as old as the modern science on Finland. As I wrote all branches of science support the same thing.

[Petalpusher]

It’s good to see that at least you and Petalpusher understand that you can’t use a k9 or k11 or k13 calculator to figure out total East Eurasian. The only thing I would suggest is that you use 3 separate West Eurasian components; ENF, WHG, Taforalt and add them for combined West Eurasian

Don't you think Tafortalt would be redondant with SSA and ENF? it's basically ENF + 1/3 SSA (maybe less WHG)

I tend to think the best low K would be

LBK
East Asia
SSA
Oceanian

Eventually add in a second run Amerindian or Siberian to sort out more specific groups in north/central Eurasian (ancient and moderns)

The old Lazaridis K6 still works well imo and is good base for reference populations peaks

https://1.bp.blogspot.com/-Oa3bBts6e3M/Ur6VH_81tpI/AAAAAAAAAUI/JGoPs1UmWyA/s1600/Lazaridis2014_EDF3_K6.png

[Komintasavalta]

It’s good to see that at least you and Petalpusher understand that you can’t use a k9 or k11 or k13 calculator to figure out total East Eurasian. The only thing I would suggest is that you use 3 separate West Eurasian components; ENF, WHG, Taforalt and add them for combined West Eurasian

I tried a bunch of different supervised ADMIXTURE runs, but I usually got really low proportions of the WHG component. In the run below, I even included EHG-mixed samples like Motala in the WHG component, but Estonians still only got only 14% of the WHG component.

I included about 30 samples as references for each component. I included at most 3 samples from each non-reference population, so that there wouldn't be too many non-reference samples relative to reference samples, and so ADMIXTURE would run faster.

I included only samples with 420,000 or more SNPs, and I excluded samples with an ".SG" suffix.

0 0 0 100 Austria_EN_LBK
0 0 0 100 Croatia_MN_Sopot
0 0 0 100 Germany_EN_LBK
0 0 0 100 Germany_EN_LBK_published
0 0 0 100 Greece_Peloponnese_N
0 0 0 100 Hungary_MN_LBK
0 0 0 100 Serbia_EN
0 0 0 100 Turkey_N
0 0 0 100 Turkey_N_published
0 0 100 0 Germany_Mesolithic
0 0 100 0 Hungary_EN_HG_Koros
0 0 100 0 Italy_North_Villabruna_HG
0 0 100 0 Latvia_HG
0 0 100 0 Lithuania_EMN_Narva
0 0 100 0 Luxembourg_Loschbour_published.DG
0 0 100 0 Romania_IronGates_Mesolithic
0 0 100 0 Serbia_IronGates_Mesolithic
0 0 100 0 Serbia_IronGates_Mesolithic_published
0 0 100 0 Spain_HG
0 0 100 0 Sweden_Motala_HG
6 0 7 87 Basque.SDG
6 1 8 86 Basque
0 7 0 93 Sardinian
7 0 6 87 Spanish_North
8 0 4 88 Spanish
5 3 1 91 Italian_North
6 2 7 85 French
8 0 4 87 Moldavian
8 0 7 84 English
8 0 6 85 French.SDG
8 0 0 91 Greek
9 0 2 90 Romanian
9 0 10 82 Icelandic
9 0 9 82 Orcadian.SDG
7 2 0 91 Maltese
7 2 9 82 Orcadian
9 0 8 83 Scottish
9 0 0 91 Sicilian
7 2 0 91 Albanian
9 0 16 75 Lithuanian
9 0 0 91 Cretan.DG
8 2 0 91 Italian_South
10 0 12 79 Norwegian.DG
5 5 0 90 Jew_Ashkenazi
10 0 4 87 Gagauz
10 0 10 80 Polish.DG
10 0 8 82 Czech
11 0 1 88 Bulgarian
11 0 6 83 Hungarian
10 1 16 73 Ukrainian_North
11 0 4 85 Croatian
11 0 8 81 Norwegian
11 0 10 78 Ukrainian
12 0 17 71 Russian
12 0 10 78 Belarusian
13 0 14 73 Estonian
13 0 0 87 Lezgin.DG
14 0 0 86 Lak
15 0 14 71 Finnish.DG
15 0 0 85 Tabasaran
15 0 0 85 Lezgin
14 2 0 84 Ingushian
15 0 0 84 Darginian
16 0 0 84 Kaitag
16 0 0 84 Adygei.SDG
16 0 0 84 Avar
16 0 0 83 Kumyk
17 0 0 83 Chechen
17 0 10 73 Mordovian
17 0 0 83 Adygei
16 1 0 83 Ossetian
18 0 13 69 Finnish
16 2 0 82 Circassian
18 0 16 65 Karelian
17 2 0 82 Balkar
19 0 13 69 Russian.SDG
19 0 13 68 Russian_Archangelsk_Krasnoborsky
19 0 22 59 Russian_Archangelsk_Pinezhsky
18 2 0 81 Kabardinian
19 0 13 68 Veps
19 1 0 80 Abazin
20 0 0 80 Karachai
25 0 12 63 Russian_Archangelsk_Leshukonsky
26 0 7 67 Tatar_Mishar
30 0 6 64 Tatar_Kazan
32 0 7 61 Chuvash
33 0 12 55 Saami.DG
34 0 4 62 Besermyan
36 0 7 57 Udmurt
33 3 0 64 Nogai_Karachay_Cherkessia
41 0 4 55 Bashkir
39 21 0 40 Nogai_Astrakhan
41 20 0 39 Nogai_Stavropol
58 29 0 13 Kalmyk
0 100 0 0 Han
0 100 0 0 Korean
100 0 0 0 Dolgan
100 0 0 0 Evenk_FarEast
100 0 0 0 Ulchi

[Sandis]

According to this admixture frequencies Estonians are more Mongoloid:

https://www.researchgate.net/publication/346952053/figure/fig5/AS:976891318136844@1609920448713/Admixture-proportions-from-three-sources-estimated-using-qpAdm-Sources-used-were.jpg

Full publication read here:

https://www.biorxiv.org/content/10.1101/2020.11.02.364521v1.full

[Arūnas]

https://www.theapricity.com/forum/images/mongol.png

[Blondie]

What's the highest amount of Mongoloid you have seen for a non-Csango/non-Szekely Hungarian?

I dont know maybe 2-3%.


How much Mong do the original Magyars have?

Thats good question, but as i know nomad magyars are closest to modern baskirs and they have:
"Fedorova's team found them to be 60.7% Caucasian and 39.3% Mongoloid. "
http://www.khazaria.com/genetics/bashkirs.html

[Zanzibar]

According to this admixture frequencies Estonians are more Mongoloid:

https://www.researchgate.net/publication/346952053/figure/fig5/AS:976891318136844@1609920448713/Admixture-proportions-from-three-sources-estimated-using-qpAdm-Sources-used-were.jpg

Full publication read here:

https://www.biorxiv.org/content/10.1101/2020.11.02.364521v1.full

That data should also includes Saami, Mansi and Nenets.

Something seems wrong about that chart though. Mari have more Neolithic admix and only slightly less Mongoloid than Udmurt? From what I remember, Mari are around 30%+ East Eurasian, but this chart show them being only a bit more than 25%. Also they should have less Neolithic than Udmurt, not more as in this Basal rich K7 data shows: https://docs.google.com/spreadsheets/d/1tFAa7oxWpcNN-OdMMjBdb4NeWKG7EkpKMzZJVW2_MME/edit#gid=2002805772

Udmurts in Basal rich K7 have around 13.24% Basal-rich (this would be the Neolithic farmer without WHG admix I believe) while the Mari are 10.37% Basal-rich. Also Udmurts have around 32.53% ANE and 19.4% East Asian. I'm sure ANE is absorbing the East Eurasian alleles because Udmurts typically score a lot higher Mongoloid than this. Meanwhile Maris in K7 have 31.34% ANE and 23.23% East Asian (am sure ANE is also absorbing the Mong score of Maris here)

Here are MDLP K16 models of Udmurts and Maris: Selkup, a Samoyedic tribe, represents the Mongoloid ancestry of Uralics, although keep mind that Selkups are not pure Mongoloid but have some West Eurasian ancestry as well. In this run, seems like a very significant Caucasoid portion of Maris is mostly Baltic-like+some Slavic-like. Also both Udmurts and Maris seem to have some Central Asian admixture (Tajik_Rushan and Uzbek) which also contains some Mongoloid ancestry as well. Anyway compare the Selkup score of Udmurt to Mari

Target: Udmurd
Distance: 249.4245% / 2.49424502
29.0 Selkup
24.0 Tadjik_Rushan
23.4 Latvian_Cesis
22.2 Finn-East
1.4 Ukrainians_north

Target: Maris
Distance: 155.1262% / 1.55126182
37.4 Latvian_Cesis
34.6 Selkup
16.0 Ukrainians_north
7.0 Uzbek
5.0 Tadjik_Rushan

Another model using G25: Udmurt are around 25% Mongoloid while Mari are close to 32% Mongoloid.

Target: Udmurt
Distance: 3.5261% / 0.03526089
47.2 RUS_Fatyanovo_BA
25.0 RUS_Krasnoyarsk_BA
21.0 Baltic_EST_BA
6.8 TJK_Sarazm_En

Target: Mari
Distance: 9.2850% / 0.09285024
31.6 RUS_Krasnoyarsk_BA
27.6 RUS_Fatyanovo_BA
23.4 RUS_Ingria_IA
13.4 Baltic_EST_BA
4.0 TJK_Sarazm_En

Also Germans, Poles, Croats are showing Mong admix (Nganasan) which seems wrong as in most other data, they don't score any East Asian.

[Komintasavalta]

Something seems wrong about that chart though. Mari have more Neolithic admix and only slightly less Mongoloid than Udmurt? From what I remember, Mari are around 30%+ East Eurasian, but this chart show them being only a bit more than 25%. Also they should have less Neolithic than Udmurt, not more as in this Basal rich K7 data shows: https://docs.google.com/spreadsheets/d/1tFAa7oxWpcNN-OdMMjBdb4NeWKG7EkpKMzZJVW2_MME/edit#gid=2002805772

It's qpAdm, and the results of qpAdm can be all over the place depending on the choice of outgroups.

Below are qpAdm models which use samples from the 1240K+HO dataset. In the first model, I used the same outgroups that were listed as outgroups in table S4 of the paper Sandis linked. The outgroups included a disproportionate number of predominantly Mongoloid populations though.

In the second model, the only change was that I added Loschbour as an outgroup. But it caused Finns to have almost 100% Yamnaya ancestry, and it reduced the Nganasan ancestry of European populations.

In the third model, I used similar outgroups as in the Dzudzuana paper by Lazaridis et al. from 2018: https://anthrogenica.com/showthread.php?14725-Joys-and-pains-of-qpAdm. Now even Poles got 16% Nganasan ancestry.

https://i.ibb.co/gv5fFhH/a.jpg

If you keep trying different combinations of outgroups long enough, you can nudge the results of qpAdm so that they resemble the results of other methods like G25. For example you can increase Mongoloid ancestry by including fewer Mongoloid outgroups. But I'm still afraid of qpAdm, because I don't what's the right method to choose the outgroups.


library(admixtools)
library(tidyverse)
library(reshape2)
library(colorspace)

left=c("Turkey_N_published","Russia_Samara_EBA_Yamnaya","Nganasan")
right=c("Han","Mbuti","Karitiana","Ulchi","Mixe")
target=c("Mansi","Chuvash","Udmurt","Iranian","Finnish","Estonian","Hungarian","Russian_Archangelsk_Krasnoborsky","Russian","Croatian","Polish.DG")

f2=f2_from_geno("path/to/v44.3_HO_public",pops=c(left,right,target))
qp=sapply(target,function(x)qpadm(f2,left,right,x) )

# qp2=apply(qp,2,function(x)x$popdrop%>%filter(feasible==T&f4rank!=0)%>%top_n(1,p)%>%select(5,7:last_col(5)))
qp2=apply(qp,2,function(x)x$popdrop%>%filter(feasible==T)%>%arrange(desc(f4rank),desc(p))%>%head(1)%>%select(5,7:last_col(5)))

qp3=do.call("rbind",qp2)
qp3$target=rownames(qp3)
qp4=melt(qp3[,-1],id.var="target")

lab=paste0(qp3$target," (",sub("e-0","e-",sub("^0","",sprintf(ifelse(qp3$p<.001,"%.0e","%.3f"),qp3$p))),")")
capt=str_wrap(paste("Outgroups:",paste(sort(right),collapse=", ")),width=76)

p=ggplot(qp4,aes(x=fct_rev(factor(target,level=qp3 $target)),y=value,fill=variable))+
geom_bar(stat="identity",width=1,position=position_fill(reverse=T),size=.0 5,color="gray20")+
geom_text(aes(label=round(100*value)),position=pos ition_stack(vjust=.5,reverse=T),size=3.5)+
labs(caption=capt)+
coord_flip()+
scale_x_discrete(expand=c(0,0),labels=rev(lab))+
scale_y_discrete(expand=c(0,0))+
scale_fill_manual(values=hex(HSV(c(30,60,300),.48, 1)))+
guides(fill=guide_legend(ncol=2,byrow=F))+
theme(
axis.text=element_text(color="black",size=11),
axis.text.x=element_blank(),
axis.ticks=element_blank(),
axis.title=element_blank(),
legend.direction="horizontal",
legend.key=element_rect(fill=NA),
legend.spacing.y=unit(.03,"in"),
legend.margin=margin(0,0,-4,0),
legend.text=element_text(size=11),
legend.title=element_blank(),
panel.border=element_rect(color="gray20",fill=NA,size=.2),
plot.caption=element_text(size=11)
)

hei=c(.05+.25*ceiling(length(unique(qp4[!is.na(qp4$value),2]))/2),.25*nrow(qp3)+.05+.18*(str_count(capt,"\n")+1))
ggdraw(plot_grid(get_legend(p),p+theme(legend.posi tion="none"),ncol=1,rel_heights=hei))
ggsave("a.png",width=6,height=sum(hei))

[Zanzibar]

It's qpAdm, and the results of qpAdm can be all over the place depending on the choice of outgroups.

Below are qpAdm models which use samples from the 1240K+HO dataset. In the first model, I used the same outgroups that were listed as outgroups in table S4 of the paper Sandis linked. The outgroups included a disproportionate number of predominantly Mongoloid populations though.

In the second model, the only change was that I added Loschbour as an outgroup. But it caused Finns to have almost 100% Yamnaya ancestry, and it reduced the Nganasan ancestry of European populations.

In the third model, I used similar outgroups as in the Dzudzuana paper by Lazaridis et al. from 2018: https://anthrogenica.com/showthread.php?14725-Joys-and-pains-of-qpAdm. Now even Poles got 16% Nganasan ancestry.

https://i.ibb.co/gv5fFhH/a.jpg

If you keep trying different combinations of outgroups long enough, you can nudge the results of qpAdm so that they resemble the results of other methods like G25. For example you can increase Mongoloid ancestry by including fewer Mongoloid outgroups. But I'm still afraid of qpAdm, because I don't what's the right method to choose the outgroups.


library(admixtools)
library(tidyverse)
library(reshape2)
library(colorspace)

left=c("Turkey_N_published","Russia_Samara_EBA_Yamnaya","Nganasan")
right=c("Han","Mbuti","Karitiana","Ulchi","Mixe")
target=c("Mansi","Chuvash","Udmurt","Iranian","Finnish","Estonian","Hungarian","Russian_Archangelsk_Krasnoborsky","Russian","Croatian","Polish.DG")

f2=f2_from_geno("path/to/v44.3_HO_public",pops=c(left,right,target))
qp=sapply(target,function(x)qpadm(f2,left,right,x) )

# qp2=apply(qp,2,function(x)x$popdrop%>%filter(feasible==T&f4rank!=0)%>%top_n(1,p)%>%select(5,7:last_col(5)))
qp2=apply(qp,2,function(x)x$popdrop%>%filter(feasible==T)%>%arrange(desc(f4rank),desc(p))%>%head(1)%>%select(5,7:last_col(5)))

qp3=do.call("rbind",qp2)
qp3$target=rownames(qp3)
qp4=melt(qp3[,-1],id.var="target")

lab=paste0(qp3$target," (",sub("e-0","e-",sub("^0","",sprintf(ifelse(qp3$p<.001,"%.0e","%.3f"),qp3$p))),")")
capt=str_wrap(paste("Outgroups:",paste(sort(right),collapse=", ")),width=76)

p=ggplot(qp4,aes(x=fct_rev(factor(target,level=qp3 $target)),y=value,fill=variable))+
geom_bar(stat="identity",width=1,position=position_fill(reverse=T),size=.0 5,color="gray20")+
geom_text(aes(label=round(100*value)),position=pos ition_stack(vjust=.5,reverse=T),size=3.5)+
labs(caption=capt)+
coord_flip()+
scale_x_discrete(expand=c(0,0),labels=rev(lab))+
scale_y_discrete(expand=c(0,0))+
scale_fill_manual(values=hex(HSV(c(30,60,300),.48, 1)))+
guides(fill=guide_legend(ncol=2,byrow=F))+
theme(
axis.text=element_text(color="black",size=11),
axis.text.x=element_blank(),
axis.ticks=element_blank(),
axis.title=element_blank(),
legend.direction="horizontal",
legend.key=element_rect(fill=NA),
legend.spacing.y=unit(.03,"in"),
legend.margin=margin(0,0,-4,0),
legend.text=element_text(size=11),
legend.title=element_blank(),
panel.border=element_rect(color="gray20",fill=NA,size=.2),
plot.caption=element_text(size=11)
)

hei=c(.05+.25*ceiling(length(unique(qp4[!is.na(qp4$value),2]))/2),.25*nrow(qp3)+.05+.18*(str_count(capt,"\n")+1))
ggdraw(plot_grid(get_legend(p),p+theme(legend.posi tion="none"),ncol=1,rel_heights=hei))
ggsave("a.png",width=6,height=sum(hei))

Can you also include the Mari and Saami in the three models you run? Wanted to see how much Yamnaya vs Turkey_N vs Nganasan they have compared to other groups..

Which of the three models do you think is the most accurate for estimating Mongoloid and Turkey_N ancestry?

[Sandis]

That data should also includes Saami, Mansi and Nenets.

Something seems wrong about that chart though. Mari have more Neolithic admix and only slightly less Mongoloid than Udmurt? From what I remember, Mari are around 30%+ East Eurasian, but this chart show them being only a bit more than 25%. Also they should have less Neolithic than Udmurt, not more as in this Basal rich K7 data shows: https://docs.google.com/spreadsheets/d/1tFAa7oxWpcNN-OdMMjBdb4NeWKG7EkpKMzZJVW2_MME/edit#gid=2002805772

Udmurts in Basal rich K7 have around 13.24% Basal-rich (this would be the Neolithic farmer without WHG admix I believe) while the Mari are 10.37% Basal-rich. Also Udmurts have around 32.53% ANE and 19.4% East Asian. I'm sure ANE is absorbing the East Eurasian alleles because Udmurts typically score a lot higher Mongoloid than this. Meanwhile Maris in K7 have 31.34% ANE and 23.23% East Asian (am sure ANE is also absorbing the Mong score of Maris here)

Here are MDLP K16 models of Udmurts and Maris: Selkup, a Samoyedic tribe, represents the Mongoloid ancestry of Uralics, although keep mind that Selkups are not pure Mongoloid but have some West Eurasian ancestry as well. In this run, seems like a very significant Caucasoid portion of Maris is mostly Baltic-like+some Slavic-like. Also both Udmurts and Maris seem to have some Central Asian admixture (Tajik_Rushan and Uzbek) which also contains some Mongoloid ancestry as well. Anyway compare the Selkup score of Udmurt to Mari

Target: Udmurd
Distance: 249.4245% / 2.49424502
29.0 Selkup
24.0 Tadjik_Rushan
23.4 Latvian_Cesis
22.2 Finn-East
1.4 Ukrainians_north

Target: Maris
Distance: 155.1262% / 1.55126182
37.4 Latvian_Cesis
34.6 Selkup
16.0 Ukrainians_north
7.0 Uzbek
5.0 Tadjik_Rushan

Another model using G25: Udmurt are around 25% Mongoloid while Mari are close to 32% Mongoloid.

Target: Udmurt
Distance: 3.5261% / 0.03526089
47.2 RUS_Fatyanovo_BA
25.0 RUS_Krasnoyarsk_BA
21.0 Baltic_EST_BA
6.8 TJK_Sarazm_En

Target: Mari
Distance: 9.2850% / 0.09285024
31.6 RUS_Krasnoyarsk_BA
27.6 RUS_Fatyanovo_BA
23.4 RUS_Ingria_IA
13.4 Baltic_EST_BA
4.0 TJK_Sarazm_En

Also Germans, Poles, Croats are showing Mong admix (Nganasan) which seems wrong as in most other data, they don't score any East Asian.

Depending on method Maris have 25-30 pct. Mongoloid.
Nganasan for Croats, Germans, Poles really seems too high.
Maris most likely were neighbours of some Eastern Baltic tribes, possibly also Iranian, like some other Finnic tribes (some extinct or assimilated). Part of their Yamnaya component here could be Baltic related.
There still is much more investigation needed to identify ancient migrations and gene flow.
Latvian_Cesis high similarity with Maris is not surprise, because Cesis has more Eastern Baltic (Lettigalian, presumably later arrivals from East) and Finnic (Livonian and Vendic) ancestry than average Latvian.

[Komintasavalta]

Which of the three models do you think is the most accurate for estimating Mongoloid and Turkey_N ancestry?

None of them. I just posted them to show how big the effect of changing outgroups is in qpAdm.

In the first model, the admixture weights are the most reasonable, but I think there are too many Mongoloid outgroups relative to Caucasoid outgroups.


Can you also include the Mari and Saami in the three models you run?

I try to restrict my scripts to samples in the 1240K+HO dataset, so it will be easy for others to rerun my scripts without having to merge datasets. The Mari samples in 1240K+HO have the suffix ".SG", and samples with that suffix often give me weird results. When I included the population named Mari.SG in the first model in my previous post, it got 61% Turkey_N, 33% Yamnaya, and 6% Nganasan.

[Zanzibar]

Depending on method Maris have 25-30 pct. Mongoloid.
Nganasan for Croats, Germans, Poles really seems too high.
Maris most likely were neighbours of some Eastern Baltic tribes, possibly also Iranian, like some other Finnic tribes (some extinct or assimilated). Part of their Yamnaya component here could be Baltic related.
There still is much more investigation needed to identify ancient migrations and gene flow.
Latvian_Cesis high similarity with Maris is not surprise, because Cesis has more Eastern Baltic (Lettigalian, presumably later arrivals from East) and Finnic (Livonian and Vendic) ancestry than average Latvian.

In some methods Maris can even score more than 30% Mongoloid up like 31-33%. I wonder if they can even get up to 35% as the upper limit.

Maris, Udmurts and Saamis are literally mixed race or almost hapas if we go by racial standards lol due to their high Mongoloid ancestry. Some Turkics like Altaians, Khakass, Kyrgyz are like the Mongoloid reverse of them as they score 65-77% Mongoloid and 23-35% Caucasoid. These Finno-Ugrics have as much Mongoloid as Castizo (Latin American who is predominantly European) have 1/4 Native American or a Quadroon (someone who is 3/4 White and 1/4 SSA).

Yes its really high which is strange cuz in most runs, Croats, Germans, Poles, etc. show noise level to none East Eurasian.

That could be it. Is the European ancestry of Finnics and other Uralics mostly Baltic-related rather than Slavic or Germanic-like? From what I have read, Maris could have Central Asian Turkic admixture as well so the Uzbek in one of the models represented that. The Iranic represented by Tajik_Rushan could indeed come from ancient Iranian tribes roaming in the area close to Volga-Ural or it could also represent Turkic affinity as many Turkic tribes were like a mix between Southern Siberian Mongoloids and Iranian tribes.

I find it crazy how high Udmurts have Iranic-like (Tajik_Rushan) admixture compare to the Mari in the MDLP K16 model. But the Iranian affinity could be either from ancient Iranian nomadic groups or from Central Asian Turks as both Maris and Udmurts seem to have a good amount of Turkic cultural and linguistic influence. Actually the Tajik_Rushan could also represent Yamnaya component as well as Tajik_Rushan and a lot of other Iranic ethnic groups score very high Steppe ancestry.

Interesting. When I use the other Latvian samples (Latvian_Dobele, Latvian) or other Baltic populations like Lithuanian, Estonian, I cannot get as good fits with the Mari as with Latvian_Cesis.

[Zanzibar]

I dont know maybe 2-3%.



Thats good question, but as i know nomad magyars are closest to modern baskirs and they have:
"Fedorova's team found them to be 60.7% Caucasian and 39.3% Mongoloid. "
http://www.khazaria.com/genetics/bashkirs.html

Oh ok. That's negligible/noise level of admixture. Seems like Csango and Szekely preserved them the best among modern Hungarians. But is there also from Turkic ancestry? I heard there were also some Turkic admixture in Hungarians?

You mean they found the Magyars to be in that range of Caucasian and Mong?

[Zoro]

It's qpAdm, and the results of qpAdm can be all over the place depending on the choice of outgroups.

Below are qpAdm models which use samples from the 1240K+HO dataset. In the first model, I used the same outgroups that were listed as outgroups in table S4 of the paper Sandis linked. The outgroups included a disproportionate number of predominantly Mongoloid populations though.

In the second model, the only change was that I added Loschbour as an outgroup. But it caused Finns to have almost 100% Yamnaya ancestry, and it reduced the Nganasan ancestry of European populations.

In the third model, I used similar outgroups as in the Dzudzuana paper by Lazaridis et al. from 2018: https://anthrogenica.com/showthread.php?14725-Joys-and-pains-of-qpAdm. Now even Poles got 16% Nganasan ancestry.

https://i.ibb.co/gv5fFhH/a.jpg

If you keep trying different combinations of outgroups long enough, you can nudge the results of qpAdm so that they resemble the results of other methods like G25. For example you can increase Mongoloid ancestry by including fewer Mongoloid outgroups. But I'm still afraid of qpAdm, because I don't what's the right method to choose the outgroups.


library(admixtools)
library(tidyverse)
library(reshape2)
library(colorspace)

left=c("Turkey_N_published","Russia_Samara_EBA_Yamnaya","Nganasan")
right=c("Han","Mbuti","Karitiana","Ulchi","Mixe")
target=c("Mansi","Chuvash","Udmurt","Iranian","Finnish","Estonian","Hungarian","Russian_Archangelsk_Krasnoborsky","Russian","Croatian","Polish.DG")

f2=f2_from_geno("path/to/v44.3_HO_public",pops=c(left,right,target))
qp=sapply(target,function(x)qpadm(f2,left,right,x) )

# qp2=apply(qp,2,function(x)x$popdrop%>%filter(feasible==T&f4rank!=0)%>%top_n(1,p)%>%select(5,7:last_col(5)))
qp2=apply(qp,2,function(x)x$popdrop%>%filter(feasible==T)%>%arrange(desc(f4rank),desc(p))%>%head(1)%>%select(5,7:last_col(5)))

qp3=do.call("rbind",qp2)
qp3$target=rownames(qp3)
qp4=melt(qp3[,-1],id.var="target")

lab=paste0(qp3$target," (",sub("e-0","e-",sub("^0","",sprintf(ifelse(qp3$p<.001,"%.0e","%.3f"),qp3$p))),")")
capt=str_wrap(paste("Outgroups:",paste(sort(right),collapse=", ")),width=76)

p=ggplot(qp4,aes(x=fct_rev(factor(target,level=qp3 $target)),y=value,fill=variable))+
geom_bar(stat="identity",width=1,position=position_fill(reverse=T),size=.0 5,color="gray20")+
geom_text(aes(label=round(100*value)),position=pos ition_stack(vjust=.5,reverse=T),size=3.5)+
labs(caption=capt)+
coord_flip()+
scale_x_discrete(expand=c(0,0),labels=rev(lab))+
scale_y_discrete(expand=c(0,0))+
scale_fill_manual(values=hex(HSV(c(30,60,300),.48, 1)))+
guides(fill=guide_legend(ncol=2,byrow=F))+
theme(
axis.text=element_text(color="black",size=11),
axis.text.x=element_blank(),
axis.ticks=element_blank(),
axis.title=element_blank(),
legend.direction="horizontal",
legend.key=element_rect(fill=NA),
legend.spacing.y=unit(.03,"in"),
legend.margin=margin(0,0,-4,0),
legend.text=element_text(size=11),
legend.title=element_blank(),
panel.border=element_rect(color="gray20",fill=NA,size=.2),
plot.caption=element_text(size=11)
)

hei=c(.05+.25*ceiling(length(unique(qp4[!is.na(qp4$value),2]))/2),.25*nrow(qp3)+.05+.18*(str_count(capt,"\n")+1))
ggdraw(plot_grid(get_legend(p),p+theme(legend.posi tion="none"),ncol=1,rel_heights=hei))
ggsave("a.png",width=6,height=sum(hei))


Yes qpAdm should only be used by experienced users who have a good understanding of how it’s used. Choice of right references definitely affect outcomes. I would use between 7 to 15 references but the key thing is they MUST be differentially related to source and target.

For example let’s focus on your Polish target. We see very different outcomes in all of your 3 scenarios. Your 3rd scenario particularly stands out because it models Poles with only 2% Yamnaya. What were the p-values and standard errors for this model ??

Even not knowing your p-values and SE, i can see your choice of outgroups/references is a poor one in scenario 3 because it violates what i just said about differentially related. You want to include a few outgroups that are much more related to Poles than Yamnaya and visa versa

[Zoro]

Don't you think Tafortalt would be redondant with SSA and ENF? it's basically ENF + 1/3 SSA (maybe less WHG)

I tend to think the best low K would be

LBK
East Asia
SSA
Oceanian

Eventually add in a second run Amerindian or Siberian to sort out more specific groups in north/central Eurasian (ancient and moderns)

The old Lazaridis K6 still works well imo and is good base for reference populations peaks

https://1.bp.blogspot.com/-Oa3bBts6e3M/Ur6VH_81tpI/AAAAAAAAAUI/JGoPs1UmWyA/s1600/Lazaridis2014_EDF3_K6.png

I don’t think so because you shouldn’t mix moderns and ancients. You also do need WHG. So if you’re doing an ancient references try not to include modern SSA because moderns are always picked by targets over corresponding ancients because they are diplois

[Komintasavalta]

In some methods Maris can even score more than 30% Mongoloid up like 31-33%. I wonder if they can even get up to 35% as the upper limit. :lightbul:

There's also about 7,000 Maris who live in XMAO (Khanty-Mansi Autonomous Okrug). I haven't seen their genetic results, but on finnougoria.ru, there was a photo of a Mari from XMAO who looked like Mansi or Khanty. Some Maris in Bashkortostan are also mixed with Bashkirs.

I think even regular Maris can be over 40% Mongoloid if you count the Mongoloid ancestry in steppe and EHG. Because VURians also have a higher EHG-WHG ratio than Finnics or Saami.


I find it crazy how high Udmurts have Iranic-like (Tajik_Rushan) admixture compare to the Mari in the MDLP K16 model. But the Iranian affinity could come from their Central Asian admixture as both Maris and Udmurts seem to have a good amount of Turkic cultural and linguistic influence. Actually the Tajik_Rushan could also represent Yamnaya component as well as Tajik_Rushan and a lot of other Iranic ethnic groups score very high Steppe ancestry.

In G25, I also noticed that if you do two-way models of Udmurts as Levänluhta plus some modern population, Tajiks give some of the best fits:

$ printf %s\\n aas\ 1F2rKEVtu8nWSm7qFhxPU6UESQNsmA-sl mas\ 1wZr-UOve0KUKo_Qbgeo27m-CQncZWb8y|while read l m;do curl "https://drive.google.com/uc?export=download&id=$m" -Lso $l;done
$ 2way2()(awk -F, 'NR==1{print"Target: "$1;for(i=2;i<=NF;i++)a[i]=$i;next}NR==2{x=$1;for(i=2;i<=NF;i++)b[i]=$i;next}{for(r=0;r<=100;r+=1){s=0;for(i=2;i<=NF;i++)s+=(a[i]-(.01*r*$i+(1-.01*r)*b[i]))^2;t[r]=s^.5};min=-1;for(i in t)if(min==-1||t[i]<min){min=t[i];minr=i};if(min!=0)printf"%f - %.0f%% %s + %.0f%% %s\n",min,minr,$1,100-minr,x}' "$@"|(read;echo "$REPLY";sort -n|awk '{$1=sprintf("%.3f",$1);sub(/^0?/,"d=")}1'))
$ 2way2 <(grep Udmurt mas) <(grep Levanluhta_IA, aas) mas|head -n17
Target: Udmurt
d=.017 - 78% Besermyan + 22% FIN_Levanluhta_IA
d=.029 - 20% Tajik_Yagnobi + 80% FIN_Levanluhta_IA
d=.030 - 22% Tajik_Rushan + 78% FIN_Levanluhta_IA
d=.030 - 23% Tajik_Shugnan + 77% FIN_Levanluhta_IA
d=.030 - 23% Tajik + 77% FIN_Levanluhta_IA
d=.031 - 21% Tajik_Ishkashim + 79% FIN_Levanluhta_IA
d=.031 - 18% Darginian + 82% FIN_Levanluhta_IA
d=.031 - 17% Avar + 83% FIN_Levanluhta_IA
d=.032 - 17% Lak + 83% FIN_Levanluhta_IA
d=.032 - 17% Tabasaran + 83% FIN_Levanluhta_IA
d=.032 - 14% Iranian_Mazandarani + 86% FIN_Levanluhta_IA
d=.032 - 17% Kubachinian + 83% FIN_Levanluhta_IA
d=.032 - 17% Kaitag + 83% FIN_Levanluhta_IA
d=.033 - 16% Chechen + 84% FIN_Levanluhta_IA
d=.033 - 17% Kalash + 83% FIN_Levanluhta_IA
d=.033 - 16% Azerbaijani_Dagestan + 84% FIN_Levanluhta_IA

[Roy]

Depending on method Maris have 25-30 pct. Mongoloid.
Nganasan for Croats, Germans, Poles really seems too high.
Maris most likely were neighbours of some Eastern Baltic tribes, possibly also Iranian, like some other Finnic tribes (some extinct or assimilated). Part of their Yamnaya component here could be Baltic related.
There still is much more investigation needed to identify ancient migrations and gene flow.
Latvian_Cesis high similarity with Maris is not surprise, because Cesis has more Eastern Baltic (Lettigalian, presumably later arrivals from East) and Finnic (Livonian and Vendic) ancestry than average Latvian.


I do wonder why many Maris look completely white to me despite their very high East / Central Asian in them. I mean 30% SSA people will nearly always look mulatto-like.

[Arūnas]

sorry Roy, you can't into Mongoloid Master Race, you are E1b :rolleyes:

[Zanzibar]

I do wonder why many Maris look completely white to me despite their very high East / Central Asian in them. I mean 30% SSA people will nearly always look mulatto-like.

It could be that Mongoloid genes are pretty weak or it could be a result of natural selection for more White-looking phenotypes? A similar case can be seen in Kazakhs and Kyrgyzs, a lot of them look pure East Asian/Siberian and don't show any Western influence in their phenotypes despite being around 25-35% Caucasoid (some Chinese Kazakhs and Kyrgyz seem to around 25% Caucasoid from DNA results that I saw). There used to a Kazakh member from Russia in this forum who is close to 40% Caucasian blood but look like a Northern Chinese or Mongol. Maris are literally the reverse versions of Kazakhs in being 30-32% East Eurasian (maybe even up to 40% as Komin suggested). Many Saamis also don't look that different from their Germanic Scandinavian counterparts despite being 25-30% Mongoloid on average (although the Saami from Kola Peninsula in Russia seems to be more Euro and less Mong due to recent Russian admixture I believe).

Negroid genes are a lot more stronger and dominant than Mongoloid in most cases. Although there are some exceptions like Ryan Giggs and Rashida Jones who looks almost completely White despite being 25-35% SSA (the latter is approx. 35% SSA from what I remember reading her classification thread on Anthroscape).

[Zanzibar]

There's also about 7,000 Maris who live in XMAO (Khanty-Mansi Autonomous Okrug). I haven't seen their genetic results, but on finnougoria.ru, there was a photo of a Mari from XMAO who looked like Mansi or Khanty. Some Maris in Bashkortostan are also mixed with Bashkirs.

I think even regular Maris can be over 40% Mongoloid if you count the Mongoloid ancestry in steppe and EHG. Because VURians also have a higher EHG-WHG ratio than Finnics or Saami.

In G25, I also noticed that if you do two-way models of Udmurts as Levänluhta plus some modern population, Tajiks give some of the best fits:

$ printf %s\\n aas\ 1F2rKEVtu8nWSm7qFhxPU6UESQNsmA-sl mas\ 1wZr-UOve0KUKo_Qbgeo27m-CQncZWb8y|while read l m;do curl "https://drive.google.com/uc?export=download&id=$m" -Lso $l;done
$ 2way2()(awk -F, 'NR==1{print"Target: "$1;for(i=2;i<=NF;i++)a[i]=$i;next}NR==2{x=$1;for(i=2;i<=NF;i++)b[i]=$i;next}{for(r=0;r<=100;r+=1){s=0;for(i=2;i<=NF;i++)s+=(a[i]-(.01*r*$i+(1-.01*r)*b[i]))^2;t[r]=s^.5};min=-1;for(i in t)if(min==-1||t[i]<min){min=t[i];minr=i};if(min!=0)printf"%f - %.0f%% %s + %.0f%% %s\n",min,minr,$1,100-minr,x}' "$@"|(read;echo "$REPLY";sort -n|awk '{$1=sprintf("%.3f",$1);sub(/^0?/,"d=")}1'))
$ 2way2 <(grep Udmurt mas) <(grep Levanluhta_IA, aas) mas|head -n17
Target: Udmurt
d=.017 - 78% Besermyan + 22% FIN_Levanluhta_IA
d=.029 - 20% Tajik_Yagnobi + 80% FIN_Levanluhta_IA
d=.030 - 22% Tajik_Rushan + 78% FIN_Levanluhta_IA
d=.030 - 23% Tajik_Shugnan + 77% FIN_Levanluhta_IA
d=.030 - 23% Tajik + 77% FIN_Levanluhta_IA
d=.031 - 21% Tajik_Ishkashim + 79% FIN_Levanluhta_IA
d=.031 - 18% Darginian + 82% FIN_Levanluhta_IA
d=.031 - 17% Avar + 83% FIN_Levanluhta_IA
d=.032 - 17% Lak + 83% FIN_Levanluhta_IA
d=.032 - 17% Tabasaran + 83% FIN_Levanluhta_IA
d=.032 - 14% Iranian_Mazandarani + 86% FIN_Levanluhta_IA
d=.032 - 17% Kubachinian + 83% FIN_Levanluhta_IA
d=.032 - 17% Kaitag + 83% FIN_Levanluhta_IA
d=.033 - 16% Chechen + 84% FIN_Levanluhta_IA
d=.033 - 17% Kalash + 83% FIN_Levanluhta_IA
d=.033 - 16% Azerbaijani_Dagestan + 84% FIN_Levanluhta_IA

Would be nice to have those Maris from XMAO uploaded to G25 as well and compared to other Maris.

Maris and other VURers are almost Hapas, you just need a bit more Mong to be 50%. :cool:

Intriguing, the Tajik score could actually be either a signal of Turkic admixture in Udmurts and other VURers or actual Steppe ancestry (Tajiks score one of the highest Yamnaya in the world after Northern/Eastern Euros) which Udmurts also derived a significant portion of their ancestries from.

[Komintasavalta]

Yes qpAdm should only be used by experienced users who have a good understanding of how it’s used. Choice of right references definitely affect outcomes. I would use between 7 to 15 references but the key thing is they MUST be differentially related to source and target.

For example let’s focus on your Polish target. We see very different outcomes in all of your 3 scenarios. Your 3rd scenario particularly stands out because it models Poles with only 2% Yamnaya. What were the p-values and standard errors for this model ??

Even not knowing your p-values and SE, i can see your choice of outgroups/references is a poor one in scenario 3 because it violates what i just said about differentially related. You want to include a few outgroups that are much more related to Poles than Yamnaya and visa versa

The p-values are shown in parentheses after the population names.

I tried adding French to outgroups because it's significantly closer to Polish than Yamnaya is, which helped reduce the Nganasan ancestry in Poles.

After that I had too high Yamnaya ancestry, so I added Kazakhstan_Mereke_MBA as an outgroup, even though it probably violates this rule: "1) It is important to realize that the answers are invalid if there has been post admixture gene-flow between left and right populations." (https://github.com/DReichLab/AdmixTools/blob/master/README.QpWave)

I tried adding Turkey_Epipaleolithic as an outgroup to reduce Turkey_N ancestry, but the percentage of Turkey_N is still 41% in Finns.

The p-values are now really low, but it might be because of a large number of outgroups: "3) We recommend keeping popright small. If large the covariance matrix of f4-statististics is likely to be poorly estimated. I then don't trust the computed p-values although the admixture weights seem to usually be reasonable." (ibid.)

https://i.ibb.co/vB0Czmn/a.png

[Zanzibar]

I do wonder why many Maris look completely white to me despite their very high East / Central Asian in them. I mean 30% SSA people will nearly always look mulatto-like.

Another case I forgot to add. Many Harnizos and Castizos from Latin America look totally Southern Euro despite having 25-35% Amerindian and even 1-5% SSA blood (many Latinos have minor Negroid dna from some African slave trade in Lat Am as well).

There is a Costa Rican member here for example, who look like any other Mediterranean Euro despite being only 62% White, the rest of his DNA is 34% Native and 4% Negroid.

So it seems that Amerindian, Mongoloid genes are weaker than Caucasian in many cases, while Negroid is stronger and more dominant than Caucasian in most cases.

[travv]

I do wonder why many Maris look completely white to me despite their very high East / Central Asian in them. I mean 30% SSA people will nearly always look mulatto-like.

Not really. Those “many” Mari were posted SCARtem and in fact they are very atypical. It’s like Turkish user post atypical Turks who can pass in Europe but in real life 99,9 % of Turks look Arab or Pakis. Same here. Most VURians don’t look white at all.

[Lemminkäinen]

QpAdm is actually a problematic method. Of course the result depends on outgroups, but qpAdm is best to use with relatively distinct and unadmixed sources. This is of course the case with all admixture, but qpAdm is misleading, because admixture percentages strongly depend on the chisq value if sources are admixed. Low chisq doesn't mean high accuracy of admixture, only a good fit. The same fit can be achieved by different admixture proportions.

[Zanzibar]

Not really. Those “many” Mari were posted SCARtem and in fact they are very atypical. It’s like Turkish user post atypical Turks who can pass in Europe but in real life 99,9 % of Turks look Arab or Pakis. Same here. Most VURians don’t look white at all.

How common are these phenos among Bashkirs, Maris and other VURers?: https://www.theapricity.com/forum/showthread.php?133699-Classify-and-compare-two-presidents-of-Bashkortostan

One of them looks pure Mongoloid while other one looks very White.

[travv]

How common are these phenos among Bashkirs, Maris and other VURers?: https://www.theapricity.com/forum/showthread.php?133699-Classify-and-compare-two-presidents-of-Bashkortostan

One of them looks pure Mongoloid while other one looks very White.

First looks normal, second looks strange to me.

[Zanzibar]

First looks normal, second looks strange to me.

Really, he looks Northern Asian. How common is pure Siberian/Mongoloid look among Bashkirs, Maris, Chuvashes and other VURers?

Do you think the second one has recent non-Bashkir ancestry?

[Ayetooey]

Estonians OFC. Lots of N up there.

[travv]

Really, he looks Northern Asian. How common is pure Siberian/Mongoloid look among Bashkirs, Maris, Chuvashes and other VURers?

Do you think the second one has recent non-Bashkir ancestry?
All Chuvash girls I posted looked 100% Mongoloid. For example these 3 Chuvash girls easily pass in Japan as native.


https://i.imgur.com/aPCHTtN.jpg
https://i.imgur.com/g85SJnj.jpg

[Zanzibar]

All Chuvash girls I posted looked 100% Mongoloid. For example these 3 Chuvash girls easily pass in Japan as native.


https://i.imgur.com/aPCHTtN.jpg
https://i.imgur.com/g85SJnj.jpg


Not really. They look too Hapa. The first president of Bashkortostan in the thread I posted can pass in NE Asia though.

[Komintasavalta]

How common is pure Siberian/Mongoloid look among Bashkirs, Maris, Chuvashes and other VURers?

A Chuvash/VUR type look like this is common among Bashkirs (these are three different persons):

https://i.imgur.com/VyY3taa.jpghttps://i.imgur.com/HslRbrj.jpghttps://i.imgur.com/nljq7vL.jpg

But I think it's impossible for a Bashkir to look fully Mongoloid.

[Sandis]

Here are another admixture results for Estonia from Bronze Age to Middle Ages:
(full study here: https://www.cell.com/fulltext/S0960-9822(19)30424-5 )

https://i.ibb.co/jH9KKvK/emss-82666-f003.jpg

EstBA – Estonian Bronze Age; EstIA – Estonian Iron Age; IngIA – Ingrian Iron Age; EstMA – Estonian Middle Ages; WHG – Western hunter-gatherers; Central MN – Central European Middle Neolithic. A. ChromoPainter/NNLS unlinked mode summarised results. B. qpAdm results

And this is from the same study for some members to understand that Y-DNA doesn't show all ancestral components (in the case of Estonians):

https://marlin-prod.literatumonline.com/cms/attachment/083161d6-5418-4b87-b243-bfaa4c80569f/fx1.jpg

[Zanzibar]

A Chuvash/VUR type look like this is common among Bashkirs (these are three different persons):

https://i.imgur.com/VyY3taa.jpghttps://i.imgur.com/HslRbrj.jpghttps://i.imgur.com/nljq7vL.jpg

But I think it's impossible for a Bashkir to look fully Mongoloid.

Wow they look identical despite being three different persons. They all look Hapa lol but gives a different feeling than the ordinary White+Asian hapa mix.

The first president of Bashkortostan (first amca in the other thread) looks almost pure Mongoloid though but I think his phenotype is pretty atypical among Bashkirs? I wager if he takes a DNA test, he is going to score more Mongoloid than most Bashkirs- probably similar to a Kazakh and Kyrgyz who are 60-70% Mongoloid. The upper Mong range for Kyrgyz seems to be 75-76% Mongoloid while for Kazakhs seems to be 70% Mongoloid (although I have seen ethnic Kazakhs from Xinjiang on G25 who score as much as 78-80% Mong but they probably have higher Mongolian ancestry than most).

[Sandis]

That could be it. Is the European ancestry of Finnics and other Uralics mostly Baltic-related rather than Slavic or Germanic-like?


From recent times also Slavic, from earlier times i think mostly Baltic-related.



From what I have read, Maris could have Central Asian Turkic admixture as well so the Uzbek in one of the models represented that. The Iranic represented by Tajik_Rushan could indeed come from ancient Iranian tribes roaming in the area close to Volga-Ural or it could also represent Turkic affinity as many Turkic tribes were like a mix between Southern Siberian Mongoloids and Iranian tribes.


Their affinity with Iranian and Turkic tribes is well documented. There is also linguistic evidence.



Interesting. When I use the other Latvian samples (Latvian_Dobele, Latvian) or other Baltic populations like Lithuanian, Estonian, I cannot get as good fits with the Mari as with Latvian_Cesis.


Latvians from Cesis looks like outliers also racially compared to a neighbour regions. It is shown in the following map. I also observed that they have more rounder traits.
Here Cesis is a mix of "Eastern Finnic" (number 2) and Western Baltic (number 1)

https://i.ibb.co/F3r79Gf/LV-racial-types.jpg

[Sandis]

Here's a 2 Estonian morphs. The left has strong Uralic traits, the right has low (n=6 for both)

https://i.ibb.co/W25LvVf/Uralid-vs-non-Uralid.png

[Zoro]

The p-values are shown in parentheses after the population names.

I tried adding French to outgroups because it's significantly closer to Polish than Yamnaya is, which helped reduce the Nganasan ancestry in Poles.

After that I had too high Yamnaya ancestry, so I added Kazakhstan_Mereke_MBA as an outgroup, even though it probably violates this rule: "1) It is important to realize that the answers are invalid if there has been post admixture gene-flow between left and right populations." (https://github.com/DReichLab/AdmixTools/blob/master/README.QpWave)

I tried adding Turkey_Epipaleolithic as an outgroup to reduce Turkey_N ancestry, but the percentage of Turkey_N is still 41% in Finns.

The p-values are now really low, but it might be because of a large number of outgroups: "3) We recommend keeping popright small. If large the covariance matrix of f4-statististics is likely to be poorly estimated. I then don't trust the computed p-values although the admixture weights seem to usually be reasonable." (ibid.)

https://i.ibb.co/vB0Czmn/a.png


Most of your p-values are indicating that your models are fails. In scenario 3 you have passing p-values but if I had to guess your standard errors are very high again rendering your models useless. Reason being is your set of outgroups is not able to properly differentiate your sources.

This explains why some of your models don’t make sense. Try using the same outgroups as Lazaridis or EurasianDNA and you’ll have better luck with qpAdm

[Zanzibar]

From recent times also Slavic, from earlier times i think mostly Baltic-related.



Their affinity with Iranian and Turkic tribes is well documented. There is also linguistic evidence.



Latvians from Cesis looks like outliers also racially compared to a neighbour regions. It is shown in the following map. I also observed that they have more rounder traits.
Here Cesis is a mix of "Eastern Finnic" (number 2) and Western Baltic (number 1)


https://i.ibb.co/F3r79Gf/LV-racial-types.jpg

I see. I also think the Western ancestry of Finno-Ugrics and other Uralics is from a Baltic-related or some other ancient Indo-European source rather than Slavic.

Is this affinity to Turkics and Iranics mostly in Mari cultural and religious norms? Yes I also heard that Mari has a lot of Turkic loanwords.

Btw what's interesting is in the same MDLP K16 model, when I decided to add Khanty, an Ugric pop, to improve the fits, the Iranic affinity of Udmurts greatly decreases, but the Slavic (Ukrainian_North) affinity great increases though but it could also be Yamnaya/Steppe ancestry from other Indo-Europeans rather than actual Slavic ancestry,

Target: Udmurd
Distance: 162.7660% / 1.62765977
40.6 Khantys
19.8 Ukrainians_north
14.6 Tadjik_Rushan
14.0 Finn-East
8.6 Latvian_Cesis
2.4 Uzbek

Target: Maris
Distance: 141.1866% / 1.41186607
31.0 Latvian_Cesis
23.8 Selkup
21.0 Ukrainians_north
14.2 Khantys
8.6 Uzbek
1.4 Tadjik_Rushan


Do Latvians from Cesis also tend to have more Mongoloid admixture than other Latvians in DNA test/studies?

[Sandis]

Is this affinity to Turkics and Iranics mostly in Mari cultural and religious norms? Yes I also heard that Mari has a lot of Turkic loanwords.


Mostly cultural and religious, but also loanwords and genetics.



Do Latvians from Cesis also tend to have more Mongoloid admixture than other Latvians in DNA test/studies?

I haven't found studies about autosomal DNA for Latvian regions, but by appearance Cesis look more Mongoloid than average, although i think that no region can have such a big Mongoloid admixture than Limbazi (North-West from Cesis).
Map should be improved, because one very characteristic phenotype from Limbazi (in the following picture) doesn't look narrow-faced at all (type 3 on the map).
From the left Uralic-influenced type from Limbazi, from the right "Eastern Finnic" type from Cesis:

https://i.ibb.co/Pg8hSmX/Limbazi-vs-Cesis.png

[Komintasavalta]

Most of your p-values are indicating that your models are fails. In scenario 3 you have passing p-values but if I had to guess your standard errors are very high again rendering your models useless. Reason being is your set of outgroups is not able to properly differentiate your sources.

This explains why some of your models don’t make sense. Try using the same outgroups as Lazaridis or EurasianDNA and you’ll have better luck with qpAdm

In my third model, I actually used a subset of the outgroups used in a paper of Lazaridis. If I would've included all the outgroups, it would've made the p-values even smaller. I was able to increase most p values to .05 or above by reducing the number of outgroups to 3, but then the admixture weights were even further from reality. Or I was also able to increase most p-values to above .05 by using 6 instead of 3 left populations. Having a high number of right populations relative to left populations seems to make the p-values smaller.

Also some of my models had low p-values because my script selected the nested model with the highest number of source populations and not the model with the highest p-value.

The first model was supposed to recreate this model from the paper linked by Sandis (https://www.biorxiv.org/content/10.1101/2020.11.02.364521v1.full), where the p-values or standard errors were not listed, but the chisq values were pretty huge:

https://i.ibb.co/MR7rvnK/20210613141653.jpg

[travv]

Not really. They look too Hapa. The first president of Bashkortostan in the thread I posted can pass in NE Asia though.

Mongoloids don’t look the same like Caucasoids don’t look the same. Even Japanese, Chinese, Koreans look different from each other. We are just as Mongolooid as Japans or Koreans are.

[Zanzibar]

Mongoloids don’t look the same like Caucasoids don’t look the same. Even Japanese, Chinese, Koreans look different from each other. We are just as Mongolooid as Japans or Koreans are.

Yes Mongs don't look the same. But VURers still overall look more like Hapas than pure Mongs. I believe you need to have more Mongoloid admixture like around Khanty, Mansi, Kazakh or Hazara levels to start looking pure Mong.

[Arūnas]

no, Chinese or Japs looks Sinid, real Mongols = VURish people

[Roy]

Yes Mongs don't look the same. But VURers still overall look more like Hapas than pure Mongs. I believe you need to have more Mongoloid admixture like around Khanty, Mansi, Kazakh or Hazara levels to start looking pure Mong.

Many Khanty people do not look fully Mongoloid.

[Roy]

Another case I forgot to add. Many Harnizos and Castizos from Latin America look totally Southern Euro despite having 25-35% Amerindian and even 1-5% SSA blood (many Latinos have minor Negroid dna from some African slave trade in Lat Am as well).

There is a Costa Rican member here for example, who look like any other Mediterranean Euro despite being only 62% White, the rest of his DNA is 34% Native and 4% Negroid.

So it seems that Amerindian, Mongoloid genes are weaker than Caucasian in many cases, while Negroid is stronger and more dominant than Caucasian in most cases.

To me a lot of those supposedly 'white' Mestizosclearly exhibit Amerindian features too. All 30% Amerindian Mexicans look already not white to me. But maybe other people are not so sensitive to spotting those elements.

[Roy]

Not really. Those “many” Mari were posted SCARtem and in fact they are very atypical. It’s like Turkish user post atypical Turks who can pass in Europe but in real life 99,9 % of Turks look Arab or Pakis. Same here. Most VURians don’t look white at all.

I have watched documentaries about Mari. It has nothing to do with some cherry-picked pictures of him that I don't even remember seeing.

[Harkonnen]

Many Khanty people do not look fully Mongoloid.

Craniofacially Khanty people cluster fully on mongoloid custer just like ancient EHG. Actually Chinese are cranially closer Euros than Khanty. Also, when the race war starts, you Poles are dead meat.

[Satem]

^^^ Wow, first saying "it's not truth there are studies" without any source, not even authors and then ad hominem, typical Harkonnen is typical

[Harkonnen]

^^^ Wow, first saying "it's not truth there are studies" without any source, not even authors and then ad hominem, typical Harkonnen is typical

I edited the wording. Hope its better now. I'm not posting the same link, I've already send it 40 times. Its very annoying to be correcting trollish posts for gazillionth time.

[Zanzibar]

Many Khanty people do not look fully Mongoloid.

Yes but how about Kazakhs, Kyrgyzs? They have the same amount of Caucasoid that Maris, many Bashkirs have of Mongoloid.

[Zanzibar]

To me a lot of those supposedly 'white' Mestizosclearly exhibit Amerindian features too. All 30% Amerindian Mexicans look already not white to me. But maybe other people are not so sensitive to spotting those elements.

I see. But can you detect the Siberian/Mongoloid phenotypical influence in many Maris?

[Harkonnen]

I see. But can you detect the Siberian/Mongoloid phenotypical influence in many Maris?

Caucasoids live in North Africa. I can assure you there are no Caucasoids in Mari-land. Roy agrees.

[SCARtem]

Not really. Those “many” Mari were posted SCARtem and in fact they are very atypical. It’s like Turkish user post atypical Turks who can pass in Europe but in real life 99,9 % of Turks look Arab or Pakis. Same here. Most VURians don’t look white at all.

You are wrong, I am placing such people not to show that all the Mari or other Volga-Ural indigenous peoples are Caucasians, but in order to destroy the stereotype that they are all Mongoloids. I am mentally closer to Caucasoid Tatars or people of mixed Tatar-Bashkir or Tatar-Mari origin and the majority. Many Caucasoid representatives of the indigenous peoples of the Volga-Volga region are descendants of the steppe Indo-Iranians (Sarmatians, Scythians, Cimmerians, Saks).
I am in favor of preserving and enhancing this Indo-Iranian heritage among the Volga-Urals. Therefore, I am in favor of pursuing a policy of eugenics among the Volga-Ural peoples, to create interethnic marriages between Caucasoid people who are descendants of Indo-Iranians.

[Zanzibar]

You are wrong, I am placing such people not to show that all the Mari or other Volga-Ural indigenous peoples are Caucasians, but in order to destroy the stereotype that they are all Mongoloids. I am mentally closer to Caucasoid Tatars or people of mixed Tatar-Bashkir or Tatar-Mari origin and the majority. Many Caucasoid representatives of the indigenous peoples of the Volga-Volga region are descendants of the steppe Indo-Iranians (Sarmatians, Scythians, Cimmerians, Saks).
I am in favor of preserving and enhancing this Indo-Iranian heritage among the Volga-Urals. Therefore, I am in favor of pursuing a policy of eugenics among the Volga-Ural peoples, to create interethnic marriages between Caucasoid people who are descendants of Indo-Iranians.

The thing is that these "Caucasoid" VURers such as Tatars, Bashkirs, Maris, etc. have very significant amounts of Mongoloid admixture as well in the range of 20-35% Siberian ancestry, which is as much as Kazakhs and Kyrgyzs have Caucasoid admixture. So even if you want to create interethnic marriages between these "Caucasoid" people, they will still get offspring and relatives who still look like Hapas or Uralic admixed.

[Sandis]

Indo-Iranians (Sycthians, Sarmatians) had a big impact not only in VUR and Central Asia, but also in the formation of Slavic tribes. There is much evidence about close interaction throughout early history.

[SCARtem]

The thing is that these "Caucasoid" VURers such as Tatars, Bashkirs, Maris, etc. have very significant amounts of Mongoloid admixture as well in the range of 20-35% Siberian ancestry, which is as much as Kazakhs and Kyrgyzs have Caucasoid admixture. So even if you want to create interethnic marriages between these "Caucasoid" people, they will still get offspring and relatives who still look like Hapas or Uralic admixed.

Firstly, I wrote not about the Volga-Urals as a whole, but only about the Caucasoid representatives, in whom this Mongoloid admixture (Huns) is minimal. I have a Tatar friend who blogs about the anthropology of the Volga Tatars, and he posts on his blog the results of genetic tests (MyTrueAncestry) of subscribers to this blog (mostly Tatars) and they all have Sarmatian, Saka, Scythian and Cimmerian genes dominating. And in racial terms, among these tested Tatars, the Pontic type dominates, just like me.

[Lemminkäinen]

I see. I also think the Western ancestry of Finno-Ugrics and other Uralics is from a Baltic-related or some other ancient Indo-European source rather than Slavic.
?

On the basis of the yDna the Estonians have Slavic ancestry, but in Finland it is negligible. The Finns have three components:

- northeastern branch of N1c1
- western branch of N1c1, common with Balts and Estonians. In a lesser amount also with East Swedes.
- branch of South Scandinavian I1

All this is old scientific data, just to mention for clarity.

Although genetic bottle necks and uniparental impact through marriages shape autosomal data, the yDna is an negligible evidence about origins of people. Language has one root, people multiple roots.

[Sandis]

On the basis of the yDna the Estonians have Slavic ancestry, but in Finland it is negligible. The Finns have three components:

- northeastern branch of N1c1
- western branch of N1c1, common with Balts and Estonians. In a lesser amount also with East Swedes.
- branch of South Scandinavian I1

All this is old scientific data, just to mention for clarity.

Although genetic bottle necks and uniparental impact through marriages shape autosomal data, the yDna is an negligible evidence about origins of people. Language has one root, people multiple roots.

Estonians are different. Slavic ancestry is more common in East, and it comes from more recent times.

https://i.ibb.co/j3P6ck0/41431-2020-699-Fig3-HTML.webp

Gray areas have no samples, i think that's because these areas are more urban and mixed (Tallinn, Narva, Maardu).

[Roy]

Estonians are different. Slavic ancestry is more common in East, and it comes from more recent times.

https://i.ibb.co/j3P6ck0/41431-2020-699-Fig3-HTML.webp

Gray areas have no samples, i think that's because these areas are more urban and mixed (Tallinn, Narva, Maardu).

I thought Estonians and Russians hardly ever mix. Apparently a lot of Russians in Estonian don't even speak Estonian.

[Roy]

I see. But can you detect the Siberian/Mongoloid phenotypical influence in many Maris?

Obviously as it is very common.

[Leto]

I thought Estonians and Russians hardly ever mix. Apparently a lot of Russians in Estonian don't even speak Estonian.
Russians/Russian speakers are mainly concentrated in Ida-Virumaa and Tallinn/Harjumaa. My uncle lives in Estonia, he's a football coach over there. People of my age and younger usually speak Estonian because it's taught from an early age and required just about everywhere in the country.

[Roy]

Russians/Russian speakers are mainly concentrated in Ida-Virumaa and Tallinn/Harjumaa. My uncle lives in Estonia, he's a football coach over there. People of my age and younger usually speak Estonian because it's taught from an early age and required just about everywhere in the country.

What Russians think of Estonians & Estonia?

[Leto]

What Russians think of Estonians & Estonia?
I'm afraid this question is hard to answer and also irrelevant to the topic. I guess they just don't think about it at all.

[Leto]

What Russians think of Estonians & Estonia?
I'm afraid this question is hard to answer and also irrelevant to the topic. I guess they just don't think about it at all.

[Lemminkäinen]

Estonian demographic history is contriversial and tricky. Those old Estonians who brought the language are not any more distinctly perceptible, but of course a part of Estonian genetics. Estonian history is episodes of
immigrations and bottle necks. What the bottle necks destroyed the immigrations strenghtened. Three bad events hit the population, the crusade time, the thirty years war time and the Sojvet time. Each of these events had a negative impact to the Estonian population and increased immigration. I am not sure how the crusades changed Estonia, but during the 17th century wars there was a big famine followed by immigrations from Latvia and Finland. The Sovjet era we know best, but let it be.

[Sandis]

I thought Estonians and Russians hardly ever mix. Apparently a lot of Russians in Estonian don't even speak Estonian.

Mixing is rare, but it exists. Old Believers created settlements near lake Peipus in 17th century, but they had less impact than Soviet occupation.

[Roy]

Mixing is rare, but it exists. Old Believers created settlements near lake Peipus in 17th century, but they had less impact than Soviet occupation.

There were Old Believers communities in Poland too in Sudovia / Podlasie, but the number of people has been way too low to have any real impact

[Zanzibar]

Obviously as it is very common.

Would you surprise if I tell you that Mari seem to be genetically closer to many Turkics/Central Asians than to most Euros?

Based on the pics that you saw, what percentage of Mari would you say pass in Poland or Northern/Eastern Europe?

[Leto]

Would you surprise if I tell you that Mari seem to be genetically closer to many Turkics/Central Asians than to most Euros?

Based on the pics that you saw, what percentage of Mari would you say pass in Poland or Northern/Eastern Europe?
Why the hell are you so focused on Russia and what does it matter how many of them would pass somewhere? You are either a Southeast Asian or a Negro from Timbuktu, so none of them will pass on your shores. I mean you keep asking so many detailed questions about Finno-Ugrics and whatnot.

[Zanzibar]

Why the hell are you so focused on Russia and what does it matter how many of them would pass somewhere? You are either a Southeast Asian or a Negro from Timbuktu, so none of them will pass on your shores. I mean you keep asking so many detailed questions about Finno-Ugrics and whatnot.

I'm mainly curious learning more about Uralics. What's wrong with that? Of course they don't pass where I am from. I'm not even interested in Russia, only some of their minorities. Why can't an outsider like me be interested about them? Anyway, I am sorry if I seem a little bit obsessed about this topic.

[Zanzibar]

Why the hell are you so focused on Russia and what does it matter how many of them would pass somewhere? You are either a Southeast Asian or a Negro from Timbuktu, so none of them will pass on your shores. I mean you keep asking so many detailed questions about Finno-Ugrics and whatnot.

I'm mainly curious learning more about Uralics. What's wrong with that? Of course they don't pass where I am from. I'm not even interested in Russia, only some of their minorities. Why can't an outsider like me be interested about them? Anyway, I am sorry if I seem a little bit obsessed about this topic.

[Dunai]

Estonians clearly preserved more Uralic genes than Hungarians, since Estonians lived in greater isolation during the many centuries and they didn't have many migratory waves as Hungary did.

[Roy]

Would you surprise if I tell you that Mari seem to be genetically closer to many Turkics/Central Asians than to most Euros?

Based on the pics that you saw, what percentage of Mari would you say pass in Poland or Northern/Eastern Europe?

This is totally expected that they are closer to Turkics.