Originally Posted by
vbnetkhio
Yes, the old averages were divided just by latitude, because it looked like only that was important for autosomal differences. For example, the old Vojvodina average from very few samples was identical to the Banija/Kordun average, so I merged them.
Now we know that's not true, so there has to be some change to the averages.
I made it like that because my Herzegovina samples are mostly from the B&H/Montenegro border with ancestry from both sides of the border. So I thought it wouldn't be correct to merge them with Montenegrins.
But from the current samples I don't see a difference between East and Old Herzegovina, so I also made the current "Montenegrin" country average which is weighted from the Serb Herzegovina and Serb South Montenegro averages.
I agree. I thought of something like this too. Bulgaria is divided into 6 averages, and I think it describes their genetics well. Serbs are spread on a wider territory so around 8 averages will be ideal.
The only things I would change, since there is already a Montenegrin country average, as I explained, I would keep the "Serb Herzegovina" and "Serb South Montenegro" as they are.
I would also keep the West and Central Serbia separate, those are large regions with millions of people, and the sample size is good.