qpAdm Neolithic admix chart for West Eurasians ANF/Natufian/CHG/Iran N/EHG/WHG etc.

[Francesco]

Just by looking at the model armenians seems a bit more plausible indeed as a donor population (yet the lowest level of Levant_N in modern Crete than in their Bronze Age Agean predecessors is a bit strange, but it could maybe be explained by some influx from continental Greece in the middle age). Don't know if Armenians could be a reasonable source populations (from a historical point of view) for Italy too

 

[eupator]

Just by looking at the model armenians seems a bit more plausible indeed as a donor population

NE Anatolians also, Caucasian Iberians and Chaldeans.

 

[Angela]

The model is interesting. One thing that seems a bit strange to me is the very low Levant Neolithic /Iran Neolithic ratio in Southern Italy, Sicily and Crete (i.e. they have high Iran N but fairly low Levant Neolithic, even lower than in Myceneans), so I wonder what donor population could increase the Iran_N component without increasing the Levant_N one: the levant bronze age populations have roughly the same propiortion between the two component, so they don't seem a viable source. Even the bronze age Anatoilan populations seem to have a more "balanced" Levant N/ Iran N ratio. Maybe it is just an issue with the sampling and not with the model.

Even more strange is that Northern Italy has 19% Iran Neo/Caucasus,and Spain not much less (15%) but no Levant Neo at all for either country. For Northern Italy, that's a ten point jump since the Republic.

What possible historical event could cause that???

Just as an aside, I love how the supposedly Nordic Romans were 65% Anatolian Neolithic, and 9% Iran Neo. So much for that theory beloved of Nazis in all time periods.

 

[Jovialis]

Something interesting I saw on Lazaridis' twitter,

PCA shows a clear shift towards Mycenaeans from the Bronze Age into the IA in Anatolia.

A shift towards the Aegean first occurs in Anatolia during the Iron Age (~1st millennium BCE) coinciding with the period of Greek colonization and the formation of the Hellenistic oikoumene (and the replacement of Anatolian languages by Greek).

https://twitter.com/iosif_lazaridis/status/1618099945608994819/photo/1

 

[Jovialis]

@IdontknowhatIamdoing

For the samples that comprise each component, which are they?

Also, I would be interested to see how your last model behaves if instead of EHG, you looked for Steppe_Samara_EBA. Then we could see non-steppe related CHG/IN.

The non-steppe related CHG/IN, in addition to Anatolian_N is considered to be a defining aspect of Mediterranean populations, as articulated in the Daunian paper. It is found even in Iron Age Latins and Etruscans, since a bit extra CHG/IN is needed to model them in addition to steppe.

 

[Power]

I might add more populations another time and change a bit the left and right pops.

The Italian_North's Natufian seems inflated here because Iran N underinflated, probably because the Standard Errors were high on that model for some reason.

Edit: i'm gonna remake this because i had to remove Jordan_PPNB from the outgroups for some pops to get a good p value but it made the model less accurate. I've found a way to fix it. For example Mycenaeans score 12.4% PPNB and with Natufian they get around 8% with the proper right pops.

This below is not fully finished yet, i might change a right pop and i will also model more people. But overall i think it's pretty good when it comes to Anatolian HG vs Natufian admix. Sadly we don't have high coverage Iran Mesolithic samples so i can make it an all mesolithic model with CHG/Iran Meso.

I merged Caucasus Hunter Gatherer Satsurblia + Iran N to estimate total CHG/Iran N related shift.

thanks for your efforts . very interesting . but i have my doubts on its accuracy

why does everybody score so much Iran_N . for example Spanish 14% , English 15% . lolwut

and why do Georgians have only 28% CHG . this cant be

 

[Idontknowwhatimdoing]

thanks for your efforts . very interesting . but i have my doubts on its accuracy

why does everybody score so much Iran_N . for example Spanish 14% , English 15% . lolwut

and why do Georgians have only 28% CHG . this cant be

We don't have samples of the CHG ancestral component of steppe, so qpAdm adds both CHG and Iran N to get something closest to it. Steppe has something CHG/Iran N related that has been in Eastern Europe for a long time/

 

[Power]

We don't have samples of the CHG ancestral component of steppe, so qpAdm adds both CHG and Iran N to get something closest to it. Steppe has something CHG/Iran N related that has been in Eastern Europe for a long time/

ok but what about the fact that Georgians are only around 28-30 % CHG ? that doesnt seem right for example

 

[Jvoquales]

I might add more populations another time and change a bit the left and right pops.

The Italian_North's Natufian seems inflated here because Iran N underinflated, probably because the Standard Errors were high on that model for some reason.

Edit: i'm gonna remake this because i had to remove Jordan_PPNB from the outgroups for some pops to get a good p value but it made the model less accurate. I've found a way to fix it. For example Mycenaeans score 12.4% PPNB and with Natufian they get around 8% with the proper right pops.

This below is not fully finished yet, i might change a right pop and i will also model more people. But overall i think it's pretty good when it comes to Anatolian HG vs Natufian admix. Sadly we don't have high coverage Iran Mesolithic samples so i can make it an all mesolithic model with CHG/Iran Meso.

I merged Caucasus Hunter Gatherer Satsurblia + Iran N to estimate total CHG/Iran N related shift.

Can you create a graph like this for Western Anatolian Greeks, Turks, especially ones from Marmara region, Thrace and other balkaners?

 

[TetaEglantina]

Already wrong because Baltic people have some Uralic/Asian input