Turkish_Balkans samples on G25

[Kaspias]

Recently new 5 Balkan Turk samples added to the Global25. For the first time, we are represented in terms of genetics with our own component.

2 From Ludogorie
1 From Northern Thrace
1 From Central and Eastern Macedonia
1 From Western Thrace

I also going to provide samples from Central Bulgaria and North Macedonia and some more from Thrace. Currently, i wanted to model available ones. (You may also want to check Thracian's thread (https://www.theapricity.com/forum/showthread.php?297215-Do-Balkan-Turks-Have-Native-Anatolian-Ancestry).)

Model1 shows Macedonian sample have some Greek admixture while others don't.

-> Abruzzo represents native Balkan.

https://i.ibb.co/HFL6Y7q/Ads-z.png

Modeling only with one Balkan sample together with Turkmen will force the algorithm to show exact non-Balkan admixture, so conqueror-migrator admixture.

https://i.ibb.co/HTPyLTg/21.png

Model3 shows Steppe/EEF/EA admixture.

https://i.ibb.co/jVcL0Zn/Model1.png

And a last ancient modeling to seperate Greek/Slav/Thracian/Turk admixtures.

https://i.ibb.co/BzhyQVs/332.png

[Marmara]

It was about time.

[Deniz]

Lol ı got the lowest Turkmen score again.

[Kaspias]

Lol ı got the lowest Turkmen score again.

Funny how genetics is not in the same parallel with phenotype.

[Ujary]

It would be better if you had used new Anatolian Turkish averages since they are the ones who immiggrated into Balkans not Turkmens. Cappadocian Greek average is not enough to model all Anatolian Turks and we don't know whether or not Turkmen average can be used for medieval Oghuzes. Moreover, Iranian_Fars average has significant East Eurasian heritage (most probably Azerbaijani effect) so it reduces the Turkmen admixture.

The best fit I have obtained is as follows:
91247

It is better than all the fits you have got as I expected and I believe this proves that modelling Balkan Turks requires Anatolian Turkish-like population rather than Turkmens.

[Kaspias]

It would be better if you had used new Anatolian Turkish averages since they are the ones who immiggrated into Balkans not Turkmens. Cappadocian Greek average is not enough to model all Anatolian Turks and we don't know whether or not Turkmen average can be used for medieval Oghuzes. Moreover, Iranian_Fars average has significant East Eurasian heritage (most probably Azerbaijani effect) so it reduces the Turkmen admixture.

The best fit I have obtained is as follows:
91247

It is better than all the fits you have got as I expected and I believe this proves that modelling Balkan Turks requires Anatolian Turkish-like population rather than Turkmens.

Thank you for contribution.

There is no single clue to make us think Balkan Turks have Anatolian background according to the results i have ever seen on both Gedmatch and Global25. It is already known they come from Anatolia, but in this route, Anatolia seems just a way for them to reach Balkans, most probably didn't mix with Anatolians. On the other hand, all Bulgarian samples from Western and Central Bulgaria despite Balkan Turks live in Eastern Bulgaria and Thrace. One expect slightly Eastern shift from natives of these regions, so Turks have that shift too. Thracian Bulgarians modeled as 50% Bulgarian + 50% North Greek. See the model:

https://i.ibb.co/QkK8ymR/Ads-z.png



Since Balkan Turks is a "bridge" between Balkan populations and Anatolian Turks, when we use Anatolian Turks instead of Turkmen it is actually expected they get lower fit. The reason why fit gets higher when we model with Turkmen is Oghuz =/= Modern Turkmen, as you said. Another anecdote is they have legit Oghuz admixture and have same migration route so do admixture, only difference is native population they have mixed. This might cause to get lower fit rather than Turkmens that heavily mixed with Iranians afterwards.

In addition, here is a comparison of their models with Thracian Bulgarian + Turkmen / Turkish_Northwest.

https://i.ibb.co/P1tLL0y/bt1.png
https://i.ibb.co/ydC65g5/bt21.png

[Kaspias]

Updated

https://i.ibb.co/7zBzLRL/Ads-z.png

[Impaler]

Updated

https://i.ibb.co/7zBzLRL/Ads-z.png

default pen or 0?

[Kaspias]

default pen or 0?

pen 0

[Impaler]

pen 0

Its seems like Global25 shift me close to Balkan_Turks. Average Romanian and Bulgarian score 0% Turkmen. And I have similar fit range as Balkan_Turkish samples.

http://i.imgur.com/16EVL2F.png (https://imgur.com/16EVL2F)

[Mingle]

Its seems like Global25 shift me close to Balkan_Turks. Average Romanian and Bulgarian score 0% Turkmen. And I have similar fit range as Balkan_Turkish samples.

http://i.imgur.com/16EVL2F.png (https://imgur.com/16EVL2F)

Model those same populations, but use these components: Bulgarian, Turkmen, Roma

[Impaler]

Model those same populations, but use these components: Bulgarian, Turkmen, Roma

I did and I added also Mortimer - Svetozar and Thracian - Balkan Turk. As you can see, Thracian score almost the same as me, just a lil bit less. I guess the Roma sample is very Balkan inflated. So, according to this model, if I have Roma ancestry, then also Thracian have which is not true. He is 100% a Balkan Turk. I am closer to Balkan_Turkish sample than Roma sample.

http://i.imgur.com/iYXDbAA.png (https://imgur.com/iYXDbAA)

[Mingle]

I did and I added also Mortimer - Svetozar and Thracian - Balkan Turk. As you can see, Thracian score almost the same as me, just a lil bit less. I guess the Roma sample is very Balkan inflated. So, according to this model, if I have Roma ancestry, then also Thracian have which is not true. He is 100% a Balkan Turk. I am closer to Balkan_Turkish sample than Roma sample.

http://i.imgur.com/iYXDbAA.png (https://imgur.com/iYXDbAA)

Well Balkan Turks are genetically closer to Romas than to Turkmens. They have little Turkic DNA. It doesn't really prove or disprove anything.

So you think you have some Turkish ancestry that makes you plot south?

[Impaler]

Well Balkan Turks are genetically closer to Romas than to Turkmens. They have little Turkic DNA. It doesn't really prove or disprove anything.

So you think you have some Turkish ancestry that makes you plot south?

Yes, that's possible. I am not sure 100%. As I said before, no one in my family know about something else. So definitely it's distant.

[Impaler]

Well Balkan Turks are genetically closer to Romas than to Turkmens. They have little Turkic DNA. It doesn't really prove or disprove anything.

So you think you have some Turkish ancestry that makes you plot south?

This is what I have in the new beta:

http://i.imgur.com/4UY5XYU.png (https://imgur.com/4UY5XYU)

http://i.imgur.com/JCmNv6l.png (https://imgur.com/JCmNv6l)

[Luke35]

Here's me, I don't know if this means anything?

"fit": 0.6053,
"Dutch": 53.33,
"Hungarian": 41.67,
"Turkmen": 3.33,
"Roma": 1.67,

"fit": 0.6065,
"Dutch": 52.5,
"Hungarian": 43.33,
"Turkmen": 4.17,

[Thracian]

Well Balkan Turks are genetically closer to Romas than to Turkmens. They have little Turkic DNA. It doesn't really prove or disprove anything.

So you think you have some Turkish ancestry that makes you plot south?

No, not really. Here is a comparison between me, Kaspias and Balkan Turks on Global 25.

https://i.ibb.co/fHHPqHM/thracian-vs-balkan-turks2.png

The fits are not suitable for some samples but it is still good to see the Turkic DNA in Balkan Turks. Adding some Northerner samples such as Romanians may cause increasing Turkmen %. I have more Iranic ancestry and it fits with the stories from my maternal side. I am usually modeled as Bulgarian/Romanian+Iranian on Gedmatch.

Impaler also has significant amount of Mongoloid. I think it is not common among Roma, they probably have non or negligible Mongoloid.

[Impaler]

Here's me, I don't know if this means anything?

"fit": 0.6053,
"Dutch": 53.33,
"Hungarian": 41.67,
"Turkmen": 3.33,
"Roma": 1.67,

"fit": 0.6065,
"Dutch": 52.5,
"Hungarian": 43.33,
"Turkmen": 4.17,

Did you use unscaled?

[Luke35]

Did you use unscaled?

Scaled, 0 pen.

Is that what you guys are doing?

[Mingle]

Yes, that's possible. I am not sure 100%. As I said before, no one in my family know about something else. So definitely it's distant.But Turkish doesn't make sense since you have more South Asian than fully Turkish people. So if you were part Turkish then it should be significantly less than what Turks have.

If it was from centuries ago (i.e. Scythian, Hunnish times) then it should be diffused throughout much of Romania rather than unique to a few individuals.

[Impaler]

Scaled, 0 pen.

Is that what you guys are doing?

Yes, pen=0.

[Kaspias]

Here's me, I don't know if this means anything?

"fit": 0.6053,
"Dutch": 53.33,
"Hungarian": 41.67,
"Turkmen": 3.33,
"Roma": 1.67,

"fit": 0.6065,
"Dutch": 52.5,
"Hungarian": 43.33,
"Turkmen": 4.17,

Your maternal site is legit, paternal side has some exotic admixture that carry both East Eurasian and South Central Asian. Guess what :wink:wink

[Impaler]

But Turkish doesn't make sense since you have more South Asian than fully Turkish people. So if you were part Turkish then it should be significantly less than what Turks have.

If it was from centuries ago (i.e. Scythian, Hunnish times) then it should be diffused throughout much of Romania rather than unique to a few individuals.

Thracian also got more South Asian than Mongoloid. We have almost same Mongoloid, around 3%, more or less. I have never seen any Roma result with my Mongoloid percentage. Also I am 20% West Asian on 23andMe and only 2% South Asian + Central Asian combined, while full Romani are even 30-35% South Asian on 23andMe with a lil bit more Western Asian. My Western Asian is 10 times higher than South Asian/Central Asian.

[Luke35]

Your maternal site is legit, paternal side has some exotic admixture that carry both East Eurasian and South Central Asian. Guess what :wink:wink

Old Magyars are :icon_whisper:?

[Impaler]

Old Magyars are :icon_whisper:?

http://i.imgur.com/wFgiNxM.jpg (https://imgur.com/wFgiNxM)

[Rgvgjhvv]

Link to the calculator my friends?

[Luke35]

Link to the calculator my friends?

http://185.144.156.77:3000/?passcode=level2p0938

[Rgvgjhvv]

http://185.144.156.77:3000/?passcode=level2p0938

Thank you! Unfortunately, I'm still very confused how this works as I don't use it much. What do I put in "Reference Pops" vs "Test Pops"?

[Luke35]

Thank you! Unfortunately, I'm still very confused how this works as I don't use it much. What do I put in "Reference Pops" vs "Test Pops"?

I put Dutch and Hungarian in reference pops as that represents my parents well. I also put in Turkmen.

I then plugged my scaled coordinates into test pops section, selected Pen=0, and pushed the magic button.

[Rgvgjhvv]

I put Dutch and Hungarian in reference pops as that represents my parents well. I also put in Turkmen.

I then plugged my scaled coordinates into test pops section, selected Pen=0, and pushed the magic button.

Ahhh yes finally I see it. Thanks a lot!

[Thracian]

Thank you! Unfortunately, I'm still very confused how this works as I don't use it much. What do I put in "Reference Pops" vs "Test Pops"?

You can pick your coordinates from Test Pops (Select Custom Users or Sponsors as Test Individuals). Put your name there and that's all. If you also want to compare yourself with other ethnicities you can select them from this title ''Select from Test Groups - Group Average will be used''.

You can select whatever you want from Reference Pops. You can select modern ethnicities such as Greeks, Italians etc... or ancient samples. It's totally up to you.

[Rgvgjhvv]

"fit": 2.2245,
"Greek": 99.17,
"Turkmen": 0.83

Hahah perfect....

[Luke35]

"fit": 2.2245,
"Greek": 99.17,
"Turkmen": 0.83

Hahah perfect....

You Greek as hell.

[Kaspias]

"fit": 2.2245,
"Greek": 99.17,
"Turkmen": 0.83

Hahah perfect....

Türk.

[Rgvgjhvv]

You Greek as hell.

https://media0.giphy.com/media/ap6wcjRyi8HoA/giphy.gif

[Rgvgjhvv]

Türk.

Truly runs DEEP in my blood

[Kaspias]

Truly runs DEEP in my blood

Deep, very deep.