what are your East Med and West Med component scores from Eurogenes K13 ?

[Vitruvius]

if i remember right, it might also additionaly be due to neoltihic levant and neolithic anatolia already both beeing largely dzudzuana derived populations, with dzudzuana beeing a mixture of basal eurasian and westeurasian. and later there was introgression from chg/neolithic iran into both populations too. there was also later neolithic levantine admixture into anatolia, but that was after the migration into europe happened. it happened before BA though so there definitely was neolithic levantine influence in BA anatolia. the thing which really differentiated early neolithic levant from early neolithic anatolia was i think around 10% ancient north african in the levant. otherwise these two populations were very similar already before anatolians moved down into the levant during bronze age.

Neolithic PPNB influence in anatolia was limited to the SE border region and not very significant. Almost all of CA/BA/IA Anatolia did not experience Levantine introgression, though it did instead experience a lot of bidirectional mixing from the Lake Van region with Proto-Armenian populations of the Armenian highlands, which is what chiefly separated post neolithic Anatolians from EEF derived populations such as Sardinians.

 

[gourd2]

Neolithic PPNB influence in anatolia was limited to the SE border region and not very significant. Almost all of CA/BA/IA Anatolia did not experience Levantine introgression, though it did instead experience a lot of bidirectional mixing from the Lake Van region with Proto-Armenian populations of the Armenian highlands, which is what chiefly separated post neolithic Anatolians from EEF derived populations such as Sardinians.

in this 2019 paper, "Late Pleistocene human genome suggests a local origin for the first farmers of central Anatolia", they find that later neolithic populations from anatolia have to be modeled with a neolithic levant source while this was not the case for the earlier neolithic anatolians.

and the later 2022 paper from Lazaridis, "Ancient DNA from Mesopotamia suggests distinct Pre-Pottery and Pottery Neolithic migrations into Anatolia", suggests the same:

"When we attempted to model Neolithic pop-ulations as mixtures of each other, we observed that at least in Anatolia (Fig. 3D), where most of the data are from and from which both Pre-Pottery and Pottery Neolithic populations have been published, an interesting distinction became clear. Pre-Pottery Neolithic populations from Central Anatolia can be modeled as mixtures of a group related to the local Pınarbaşı Epipaleolithic with variable (~30 to 70%) Mesopotamian admixture, suggesting that Pre-Pottery cultures of Anatolia may have been formed with the contribution of both local hunter-gatherers and migrants from the east, where agriculture first appeared. But we can not model the Pottery Neolithic Anatolians with just these two sources and instead require an extra ~6 to 23% Levantine Neolithic admixture. "

and i believe the pottery neolithic samples in this study are also from Barcin, which is located in northwestern turkey. so it seems neolithic levantine influence in late neolithic anatolia was not limited to southeastern Anatolia.

 

[Power]

regarding the levant thing : just look at your Natufian scores and you are set

 

[davef]

When I said ancient Anatolian exists in Greece and Italy I meant BA not Neolithic (the latter is in every euro and middle eastern group).

 

[Power]

When I said ancient Anatolian exists in Greece and Italy I meant BA not Neolithic (the latter is in every euro and middle eastern group).

the question is how the Anatolian BA would get broken down with a neolithic model . the important score when it comes to the levant question is the natufian .

 

[italouruguayan]

Mine....

North_Atlantic	27.06%
Baltic	10.41%
West_Med	15.74%
West_Asian	2.66%
East_Med	17.88%
Red_Sea	2.99%
South_Asian	-
East_Asian	-
Siberian	0.66%
Amerindian	18.62%
Oceanian	-
Northeast_African	0.83%
Sub-Saharan	3.13%

[th]

Population​

[/th][th]

Percentage​

[/th]​

 

[Vitruvius]

in this 2019 paper, "Late Pleistocene human genome suggests a local origin for the first farmers of central Anatolia", they find that later neolithic populations from anatolia have to be modeled with a neolithic levant source while this was not the case for the earlier neolithic anatolians.

and the later 2022 paper from Lazaridis, "Ancient DNA from Mesopotamia suggests distinct Pre-Pottery and Pottery Neolithic migrations into Anatolia", suggests the same:

"When we attempted to model Neolithic pop-ulations as mixtures of each other, we observed that at least in Anatolia (Fig. 3D), where most of the data are from and from which both Pre-Pottery and Pottery Neolithic populations have been published, an interesting distinction became clear. Pre-Pottery Neolithic populations from Central Anatolia can be modeled as mixtures of a group related to the local Pınarbaşı Epipaleolithic with variable (~30 to 70%) Mesopotamian admixture, suggesting that Pre-Pottery cultures of Anatolia may have been formed with the contribution of both local hunter-gatherers and migrants from the east, where agriculture first appeared. But we can not model the Pottery Neolithic Anatolians with just these two sources and instead require an extra ~6 to 23% Levantine Neolithic admixture. "

and i believe the pottery neolithic samples in this study are also from Barcin, which is located in northwestern turkey. so it seems neolithic levantine influence in late neolithic anatolia was not limited to southeastern Anatolia.

Using Pinarbasi - a single sample from 14,000 BC in a mixed chronology source model to infer what "pure" neolithic anatolian ancestry should look like against then contemporary or near contemporary neolithic levantines and mesopotamians while excluding neolithic caucasian references is simply poor modelling. No reasonable person can or should take it seriously.

It's extremely clear on our prehistoric PCA that neolithic anatolians were experiencing a minor shift towards caucasian populations which drastically accelerated at the turn of the copper age. Meanwhile, levantine populations showed influence of being drastically affected by Anatolian population introgression in the neolithic and also later Caucasian and Iranian rich sources during the bronze age. As for what Anatolians looked like prior to the neolithic, we scarcely have a clue and Pinarbasi is only a single data point which is very chronologically removed from all later samples of the peninsula. Despite this, Pinarbasi still clusters much more comparably close to neolithic anatolians than Natufians to neolithic levantines, which says a good deal about the direction of ancestry flow across this period.

 

[gourd2]

the question is how the Anatolian BA would get broken down with a neolithic model . the important score when it comes to the levant question is the natufian .

considering what these papers found, such a model would have to exclude anatolian late neolithic populations from the sources and keep levant neolithic.

 

[Stuvanè]

Mine

Population

North_Atlantic

East_Med

West_Med

Baltic

West_Asian

Red_Sea

Siberian

Northeast_African

[td]

Percent​

[/td]​

[td]

1​

[/td]​

[td]

28.32​

[/td]​

[td]

2​

[/td]​

[td]

22.12​

[/td]​

[td]

3​

[/td]​

[td]

21.39​

[/td]​

[td]

4​

[/td]​

[td]

13.52​

[/td]​

[td]

5​

[/td]​

[td]

9.68​

[/td]​

[td]

6​

[/td]​

[td]

4.44​

[/td]​

[td]

7​

[/td]​

[td]

0.29​

[/td]​

[td]

8​

[/td]​

[td]

0.24
​

[/td]​

[​

 

[gourd2]

Using Pinarbasi - a single sample from 14,000 BC in a mixed chronology source model to infer what "pure" neolithic anatolian ancestry should look like against then contemporary or near contemporary neolithic levantines and mesopotamians while excluding neolithic caucasian references is simply poor modelling. No reasonable person can or should take it seriously.

It's extremely clear on our prehistoric PCA that neolithic anatolians were experiencing a minor shift towards caucasian populations which drastically accelerated at the turn of the copper age. Meanwhile, levantine populations showed influence of being drastically affected by Anatolian population introgression in the neolithic and also later Caucasian and Iranian rich sources during the bronze age. As for what Anatolians looked like prior to the neolithic, we scarcely have a clue and Pinarbasi is only a single data point which is very chronologically removed from all later samples of the peninsula. Despite this, Pinarbasi still clusters much more comparably close to neolithic anatolians than Natufians to neolithic levantines, which says a good deal about the direction of ancestry flow across this period.

PCA clustering can't be used for ancestry inference.

Also, they didn't exclude neolithic caucasus/iranian references. From the 2019 paper:
"Ancient Iran/Caucasus populations and contemporary South Asians do not share more alleles with ACF (|D| < 1.3 SE). Likewise, qpAdm modeling suggests that the AAF gene pool still constitutes more than 3/4 of the ancestry of ACF 2000 years later (78.7 ± 3.5%; Supplementary Tables 4 and 7) with additional ancestry well modeled by the Neolithic Levantines (χ2p = 0.115) but not by the Neolithic Iranians (χ2p = 0.076; the model estimated infeasible negative mixture proportions) (Supplementary Tables 4 and 7)."

In the Lazaridis 2022 paper they included CHG in the right outgroup set and mesopotamian neolithic already includes this type of ancestry.

and most importantly explain this: why is it possible to model earlier central anatolian populations without neolithic levant, but later anatolian populations from central and even north western anatolia require neolithic levant or else the same models fail? there must have been a change, and it didn't come from a mesopotamian/caucasus direction.

 

[davef]

I never tested but had Gemini guess what I should be in this k13

• North_Atlantic: 27.12%
• East_Med: 25.75%
• West_Med: 19.38%
• West_Asian: 11.38%
• Baltic: 10.12%
• Red_Sea: 4.62%
• South_Asian: 0.62%
• Northeast_African: 0.53%
• Sub-Saharan: 0.25%
• Amerindian: 0.23%
• Siberian: 0.00%
• East_Asian: 0.00%

Few things might be off by couple points but it’s close enough and still very likely

 

[Vitruvius]

PCA clustering can't be used for ancestry inference.

Also, they didn't exclude neolithic caucasus/iranian references.

That's false. From your quote claiming 6-23% Levantine PPNB ancestry the model which is referenced is Figure D. It is a simple forced 3 way admixture model with no Caucasian nor Iranian source populations as can be clearly seen. It uses Pinarbasi and Levant PPN despite the fact that these samples are 6000 years removed from one another in addition to the fact that it's already well known that PPNB exhibits large influence from Pinarbasi-like Anatolian populations. It also includes Cyprus at the upper end of that 23% figure which is not even part of Anatolia at all.

"In the Lazaridis 2022 paper they included CHG in the right outgroup set and mesopotamian neolithic already includes this type of ancestry."

CHG (Also a very archaic population) is nothing like Mesopotamian and Mesopotamian (Neolithic northern Mesopotamian in reality) is also not a suitable proxy for neolithic Armenian ancestry, which did in fact have a very large and dominating influence in the Anatolian Bronze age, unlike PPNB which instead averages 1% across the entire BA anatolian population. This influence between BA Anatolia and the Caucasus is similarly reflected in a highly shared material culture and trade network that has been long since recognized.

Looking at this figure, one should also keep in mind that this is in addition to the fact that Levant PPNB is roughly half Anatolian derived in of itself and not similar or particularly close to the more archaic Natufian. We do not see the same phenomenon with Pinarbasi in relation to neolithic Anatolians who, despite only having but one sample, shows much closer relation and continuity comparatively. During the bronze age, Levant PPNB ancestry within the borders of Turkey (which is of course much larger than the geographic border of the Anatolian peninsula) is instead limited to the Turkish Levant, the Anatolian/Levant border region and the Turkish border with the area of Mesopotamia/Iraq. This can be seen quite clearly in the samples below which are derived from such areas in Turkey by comparison with the same model.

 

[Power]

considering what these papers found, such a model would have to exclude anatolian late neolithic populations from the sources and keep levant neolithic.

levant neolithic is roughly half ANF half Natufian . i never use levant neolithic

 

[davef]

levant neolithic is roughly half ANF half Natufian . i never use levant neolithic

Yeah less overlap between samples the better! Else one would inflate higher than reality

 

[gourd2]

That's false. From your quote claiming 6-23% Levantine PPNB ancestry the model which is referenced is Figure D. It is a simple forced 3 way admixture model with no Caucasian nor Iranian source populations as can be clearly seen. It uses Pinarbasi and Levant PPN despite the fact that these samples are 6000 years removed from one another in addition to the fact that it's already well known that PPNB exhibits large influence from Pinarbasi-like Anatolian populations. It also includes Cyprus at the upper end of that 23% figure which is not even part of Anatolia at all.

it doesn't really matter how far apart Pinarbasi and neolithic levant are chronologically. fact is, that the earlier neolithic anatolians can be modeled as a mixture of Pinarbasi/Iran_N in the 2019 paper or Pinarbasi/Mesopotamian in the 2022 paper but this doesn't work anymore for the later neolithic anatolians. instead the models require neolithic levant. you have to explain this discrepancy somehow.

in your models you are using Turkey_N which includes those admixed later neolithic populations. if you want to find the admixture proportions of neolithic levant in late neolithic anatolia why are you including late neolithic anatolia in your sources?
swap it out with an earlier neoltihic population like Boncuklu and you get something like this for late neolithic/chalcolithic

and for bronze age

and i think most of these bronze age samples are from central/western anatolia.

you can also use Natufian instead of neolithic levant

i admit that using only Boncuklu_N is not optimal, however if you want to argue against neolithic levant introgression into late neolithic anatolia you absolutely can't use late neolithic anatolian samples in your source populations.

 

[Power]

^ it doesnt make sense to use levant neolithic . as i said levant neolithic is roughly half anatolian neolithic itself

 

[gourd2]

^ it doesnt make sense to use levant neolithic . as i said levant neolithic is roughly half anatolian neolithic itself

you can also use Natufian if you want, i updated the post with a table including Natufian as source. in the end all of these populations have a certain overlap. there was also a cline from northwestern anatolia towards southern levant already existing in early neolithic. so the observed shift in central anatolia might not have come from 6-23% from southern levant directly but from a more northern levantine or southeastern anatolian population with more affinity to southern levantines/natufians. in that case the contribution would probably be higher than 6-23%. in any case there probably was a movement from southeast to northwest during the neolithic in anatolia.