Azeri DNA

[Kyp]

Version 1:
Calculator used: Dodecad k12b
Samples:
Anatolian: Armenian, Cappadocians combined
Caucasus: Abkhazians,
Arab: Bedouins
Iranian: Medieval Iranians from Iran and Kurdish ares combined
Mongol: Mongolians
Turkic: Turkmens from Uzbekistan


any idea how I can improve it?

Note: Iranian Azeris have by far the highest sample size. Turkmen_Uzbekistan eats additional Iranic ancestry.

[Leto]

Don't mix ancients with modern ones. Also, I would use Lezgins or at least Georgians instead of Abkhazians.

Kurd
Armenian
Lezgin
Karakalpak
Lebanese

[Leto]

Target: Azeri_Turkey
Distance: 2.1210% / 0.02121004
52.8 Armenian
28.6 Kurdish
10.2 Karakalpak
8.4 Tabasaran

Target: Azeri_Dagestan
Distance: 1.0911% / 0.01091129
38.8 Tabasaran
29.6 Kurdish
24.8 Armenian
6.8 Karakalpak

Target: Azeri
Distance: 0.8776% / 0.00877632
53.4 Kurdish
32.6 Armenian
11.0 Karakalpak
3.0 Tabasaran

Target: Azeri_Turkey
Distance: 2.1874% / 0.02187441 | ADC: 0.25x
43.6 Kurdish
42.2 Armenian
8.6 Karakalpak
5.6 Tabasaran

Target: Azeri_Dagestan
Distance: 1.2004% / 0.01200412 | ADC: 0.25x
42.8 Kurdish
38.2 Tabasaran
13.8 Armenian
5.2 Karakalpak

Target: Azeri
Distance: 1.0525% / 0.01052496 | ADC: 0.25x
70.2 Kurdish
20.4 Armenian
9.4 Karakalpak

[Kyp]

Don't mix ancients with modern ones. Also, I would use Lezgins or at least Georgians instead of Abkhazians.

Kurd
Armenian
Lezgin
Karakalpak
Lebanese

What is the relation with Lebanese though? My goal is not to make the distances as tight as possible. It should make sense from an historic point of view in the most simple way possible

g25 is too accurate for reality in my opinion too.

[Leto]

What is the relation with Lebanese though?

And why Bedouin? I used the Lebs as a purely Near Eastern population. And none of the Azeri references picked it up.

[Leto]

Azeris are pretty much a mix of Northwest Iranic people like Kurds, North Caucasians like Lezgins and a Turkic population like Karakalpaks plus maybe some Armenian/Anatolian mixed in there.

[Thracian]

Good idea and seems good. However, I also agree with Leto, you should use ancient and modern samples separately.

I personally would separate regions like below,

-Caucasus, Georgians
-Eastern Anatolia and Mesopotamian, Armenian and Assyrian
-Northeast Anatolian, Armenian Hemshin and Greek trabzon
-Iranic, would exclude Bandari also add Kurds and Talysh
-Turkic, Turkmen Uzbekistan

[Kyp]

And why Bedouin? I used the Lebs as a purely Near Eastern population. And none of the Azeri references picked it up.

Abbasids etc.? I could integrate a Levantine component too. Aren't there more pure Levantine groups than Lebanese?

Azeris are pretty much a mix of Northwest Iranic people like Kurds, North Caucasians like Lezgins and a Turkic population like Karakalpaks plus maybe some Armenian/Anatolian mixed in there.

I used Anatolian because it is known that Shia Turks from Anatolia went to Azerbaijan in Ottoman times.

[Kyp]

Good idea and seems good. However, I also agree with Leto, you should use ancient and modern samples separately.

I personally would separate regions like below,

-Caucasus, Georgians
-Eastern Anatolia and Mesopotamian, Armenian and Assyrian
-Northeast Anatolian, Armenian Hemshin and Greek trabzon
-Iranic, would exclude Bandari also add Kurds and Talysh
-Turkic, Turkmen Uzbekistan

I used Medieval Iranian and Kurds because they lack Turkic/Mongol DNA, other than that they're almost identical to modern ones but I understand the issue..


I will take yours and Leto's advice into consideration thanks

[porpozontokonto]

why are you using uzbekstani turkmen samples as basis for turkic? just curious