How can Turkish people from Western Anatolia be closer to Iranians than they are to most Europeans?

[reboun]

Another crazy thing is Lebanese Christians being closer to Greeks than Anatolian Turks while not being at Greece's doorstep.

Southern Siberian admixture among Turkish people creates the difference in my opinion.

[Petalpusher]

I would say it's from two sources. One coming from Levant_BA people and one from BA Anatolia, but this admixture event likely happend during the Early Roman era. Both populations had around 30% Iran_N/CHG without any Steppe. Even North Italians have significant amount of post-Neolithic admixture from Anatolia/Levant.

I ve discussed about this, several years ago but difficult around here. Yes it's possible even N.Italians have a bit of it. It's just that mathematically, it's impossible anything below Tuscans, doesn't have it. You basically make a triangle between mesolithic, earliy neo and steppe and whatever doesn't fall into it, has to have something else. Most Europe conforms to this simple model without the need of additionnal admixture, even Iberians, Sardinians, etc.. so it's already very conservative. Then indeed, it can be a lot of things, either post BA or "different BA" at different times.

[Aren]

I ve discussed about this, several years ago but difficult around here. Yes it's possible even N.Italians have a bit of it. It's just that mathematically, it's impossible anything below Tuscans, doesn't have it. You basically make a triangle between mesolithic, earliy neo and steppe and whatever doesn't fall into it, has to have something else. Most Europe conforms to this simple model without the need of additionnal admixture, even Iberians, Sardinians, etc.. so it's already very conservative. Then indeed, it can be a lot of things, either post BA or "different BA" at different times.

It looks like it when you model modern Europeans with the available Anatolian Neolithic samples and EBA Steppe, yes. But IMO this is not the most accurate method. Neolithic Barcin was not the exact source for most of the European farmers, especially not for the Cardial Ware since Barcin seem to have some actual Iran/Caucasus/Levantine admixture lacking in European farmers(outside of Greece). Second mistake most people do is to use Yamnaya to model modern Europeans. Yamnaya with it's R-Z2103 was not ancestral to most of Europe, it's only relevant for Greeks, Albanians and non-Iranian West Asians. Yamnaya has too much CHG and too little EHG, so populations who have some extra non-Steppe CHG will have it eaten up by Yamnaya. Relevant sources would be Corded Ware and Bell Beakers.
But most importantly if we want to find out whether a population has post-Neolithic admixture from outside of Europe we need to use post-Neolithic samples. And we have lots of MN/CA farmers from all over Europe to use.

Here's a simple run for modern Northern Italians that showcases this.

Target: Italian_Lombardy
Distance: 1.3483% / 0.01348316
43.4 ITA_Grotta_Continenza_CA
38.8 Corded_Ware_DEU
12.8 TUR_Kaman-Kalehoyuk_MLBA
5.0 Levant_Sidon_MBA

Without Anatolia/Levant_BA the fit is a lot worse, you simply need something with Natufian/CHG/Iran admixture without Steppe.

Target: Italian_Lombardy
Distance: 2.8206% / 0.02820605
56.4 ITA_Grotta_Continenza_CA
43.6 Corded_Ware_DEU

[Petalpusher]

It looks like it when you model modern Europeans with the available Anatolian Neolithic samples and EBA Steppe, yes. But IMO this is not the most accurate method. Neolithic Barcin was not the exact source for most of the European farmers, especially not for the Cardial Ware since Barcin seem to have some actual Iran/Caucasus/Levantine admixture lacking in European farmers(outside of Greece). Second mistake most people do is to use Yamnaya to model modern Europeans. Yamnaya with it's R-Z2103 was not ancestral to most of Europe, it's only relevant for Greeks, Albanians and non-Iranian West Asians. Yamnaya has too much CHG and too little EHG, so populations who have some extra non-Steppe CHG will have it eaten up by Yamnaya. Relevant sources would be Corded Ware and Bell Beakers.
But most importantly if we want to find out whether a population has post-Neolithic admixture from outside of Europe we need to use post-Neolithic samples. And we have lots of MN/CA farmers from all over Europe to use.

Here's a simple run for modern Northern Italians that showcases this.

Target: Italian_Lombardy
Distance: 1.3483% / 0.01348316
43.4 ITA_Grotta_Continenza_CA
38.8 Corded_Ware_DEU
12.8 TUR_Kaman-Kalehoyuk_MLBA
5.0 Levant_Sidon_MBA

Without Anatolia/Levant_BA the fit is a lot worse, you simply need something with Natufian/CHG/Iran admixture without Steppe.

Target: Italian_Lombardy
Distance: 2.8206% / 0.02820605
56.4 ITA_Grotta_Continenza_CA
43.6 Corded_Ware_DEU

Yeah i was talking about a time CHG wasn't even released (even less Iran_N and newer samples) when i noticed it, got a lot of crap back then for pointing it out... But i agree Yamnaya is not the best direct source, in the same way some models with Samara give some pop that are very unlikely to have some relevant levels of Steppe. Because they have high CHG/Iran_N, and there s some good ANE in this to begin with, a bit of WHG and that's enough to confuse it for Steppe since it looks like some ANE+WHG (so EHG like) and basal. However i think BB and CWC can produce their own admixture artefacts as they are IE acquiring LN/CA europe type of admix.

Is it G25 btw and which dataset?

[Aren]

Yeah i was talking about a time CHG wasn't even released (even less Iran_N and newer samples) when i noticed it, got a lot of crap back then for pointing it out... But i agree Yamnaya is not the best direct source, in the same way some models with Samara give some pop that are very unlikely to have some relevant levels of Steppe. Because they have high CHG/Iran_N, and there s some good ANE in this to begin with, a bit of WHG and that's enough to confuse it for Steppe since it looks like some ANE+WHG (so EHG like) and basal. However i think BB and CWC can produce their own admixture artefacts as they are IE acquiring LN/CA europe type of admix.

Is it G25 btw and which dataset?

Yes CWC and BB both have some substantial European_MN/CA, but the point was not to show actual Steppe input. Using early Corded Ware is a better choice for that.

Yes G25, scaled. I always use scaled coordinates.
Ancient samples:https://drive.google.com/file/d/1F2r...SQNsmA-sl/view
Modern:https://drive.google.com/file/d/1wZr...CQncZWb8y/view

[Defcon2]

Target: Defcon2_scaled
Distance: 4.3510% / 0.04351024
51.8 ITA_Grotta_Continenza_CA
41.8 Corded_Ware_DEU
6.4 Levant_Sidon_MBA

Target: Defcon2_scaled
Distance: 2.9326% / 0.02932553
51.2 ITA_Grotta_Continenza_CA
41.6 Corded_Ware_DEU
7.2 MAR_Taforalt

Sorry, I couldn't resist...

[Aren]

Target: Defcon2_scaled
Distance: 4.3510% / 0.04351024
51.8 ITA_Grotta_Continenza_CA
41.8 Corded_Ware_DEU
6.4 Levant_Sidon_MBA

Target: Defcon2_scaled
Distance: 2.9326% / 0.02932553
51.2 ITA_Grotta_Continenza_CA
41.6 Corded_Ware_DEU
7.2 MAR_Taforalt

Sorry, I couldn't resist...

You need a farmer source with more WHG.

Something like this
Iberia_Central_CA,0.1263435,0.1690857,0.0548712,-0.029716,0.095864,-0.0212652,-0.0072262,0.0002305,0.0659078,0.0904348,-0.007064,0.0140125,-0.0315535,-0.0146222,-0.0047842,0.0045082,0.0117018,0.0010768,-0.0014452,-0.0030325,0.01541,0.0018858,-0.0118318,-0.030607,0.0013172

[Defcon2]

You need a farmer source with more WHG.

Something like this
Iberia_Central_CA,0.1263435,0.1690857,0.0548712,-0.029716,0.095864,-0.0212652,-0.0072262,0.0002305,0.0659078,0.0904348,-0.007064,0.0140125,-0.0315535,-0.0146222,-0.0047842,0.0045082,0.0117018,0.0010768,-0.0014452,-0.0030325,0.01541,0.0018858,-0.0118318,-0.030607,0.0013172

Target: Defcon2_scaled
Distance: 3.4515% / 0.03451455
43.4 Iberia_Central_CA
33.4 Corded_Ware_DEU
17.4 TUR_Kaman-Kalehoyuk_MLBA
5.8 MAR_Taforalt

[Aren]

Target: Defcon2_scaled
Distance: 3.4515% / 0.03451455
43.4 Iberia_Central_CA
33.4 Corded_Ware_DEU
17.4 TUR_Kaman-Kalehoyuk_MLBA
5.8 MAR_Taforalt

Use both farmer populations.

[Defcon2]

Use both farmer populations.

Target: Defcon2_scaled
Distance: 2.9293% / 0.02929282
49.2 ITA_Grotta_Continenza_CA
41.4 Corded_Ware_DEU
7.2 MAR_Taforalt
2.2 Iberia_Central_CA