EHG in South Caucasian farmers?

[Aramu]

I was trying to reproduce Lazaridis 2022 Southern Arc paper models with qpadm when I realized that there is a persistent EHG in South Caucasian farmers. It is low level, and I changed various right pops but it did not disappear.
What You think about it? Do I do something wrong with right pops or its a real thing. If it is real can we expect that in Upper Paleolithic there was EHG like ancestry in Caspian basin similar to WHG in Dzudzuana in Black sea basin. Or its a recent event.

   
   
Then in Late Chalcholithic the EHG increase as expected in samples from Areni cave. This increase is related  to Soyuq Bulag kurgans. Which appear in NW Iran also, in Se Girdan.
   

I use Rstudio
Right pops used
right=c('Mbuti.DG', 'Russia_Kostenki14.SG', 'Russia_MA1_HG.SG', 'Italy_North_Villabruna_HG', 'Russia_Sidelkino_HG.SG', 'China_Tianyuan', 'Czech_Vestonice16', 'Ethiopia_4500BP.DG', 'Russia_Ust_Ishim.DG', 'Belgium_UP_GoyetQ116_1','Switzerland_Bichon.SG','Russia_LakeBaikal_N.SG','Iran_ShahrISokhta_BA2','Georgia_Satsurblia.SG','Turkey_Boncuklu_N','Russia_Samara_EBA_Yamnaya','Indian_GreatAndaman_100BP.SG','Morocco_Iberomaurusian')

[Kale]

Shouldn't be terribly controversial to suggest trace level EHG in South-Caucasian farmers, after all...
https://genarchivist.com/showthread.php?...71#pid6071
I would advise against using SIS_BA2 as a right pop in this particular setup though, it's much younger and could violate the rule of 'no post-admixture interaction between left and right pops'

On the other hand... MasisBlur, Azerbaijan_LN, and HajjiFiruz_CA can all get passing models as 100% Mesopotamia_PPN.

[Aramu]

I modeled Iraq PPNA with same left and right pops and it do not have any EHG.
Apparently the extra EHG was present in South Caucasus even prior farmers came.