@Angela
I think DNA Land called it "South European". Balkan is Albanian + Bulgarian + Greek, and South/Central European is Bergamo + Tuscan, isn't that right?
@Sile
51%!? It's really hard to believe in such number, especially if we consider DNA Land is based on modern people. Besides, don't you think odd a Venetian being closer to Balkan than to Bergamo? Likely the link you mentioned helps to explain the result, but it would be still a bad result, imo, even if Bessica/Bessega (where a grandfather of my mother had roots) has been colonized by "Bessoi" soldiers in Roman times.
Btw, the Balkan cluster exists also in 23andMe, and my mother gets between 7 and 8% there (version 4). Well, it seems to me there are sufficient evidences (see also previous posts) showing DNA Land "have much room for improvement", as davef says.
I'm sure you know better than I do. I don't pay very much attention to these things any more. Really, it's only ancient dna and how it proves or disproves archaeology and ancient history that interests me.
What I do know is that on Kurd's calculator with no Bergamo or Tuscan reference samples, I always come out as someone from the Balkans, usually Bulgarian or Albanian or something in the top two.
Cavalli-Sforza pointed out forty years ago that northern Italians are close to the people of the northern Balkans (not Greeks). That's old hat. However, a calculator that tells me I'm Bulgarian is useless because I'm not. On any calculator with Bergamo and Tuscan samples I'm right between them. When the "North Italy" sample from Piemonte is included, I'm closest to them. Obviously, calculators with those reference samples are better at "placing" me.
This dividing Europe into "clusters" is fraught with difficulty, especially in southern Europe. DNA Land has obviously aggregated those alleles which northern Italians
share with Balkan peoples, and includes them in the Balkans cluster or labels them as "Balkan". If they created a "Northern Italian" cluster of those shared north Italian/Balkan genes and called them north Italian, people in the Balkans would get big North Italian percentages. One isn't more "right" than another.
What I do know is that northern Italians are not Bulgarians, and Tuscans are not Albanians, even if they might place relatively near each other on a PCA plot or they share some alleles. Yes, they were equally impacted by certain population migrations, but on the other hand, there are definite differences in terms of which "northern" or central European groups affected them, as just one example. Northern Italians got more from the "Celts" and the "Germanics", although the Balkans got some. On the other hand, the Balkans got more influence from the "Slavs", and even in some cases from the Central Asiatics. So, they're not the same people, and any calculator that gives that impression isn't any good.
Obviously, we know that someone from the Veneto should be closer to someone from Bergamo or Brescia than to someone from Bulgaria.
Oh, Genographic is another example. Do they still cluster Italians and Greeks together? That's bound to give a whole different set of percentages.
No, I don't think these programs and calculators are very useful other than for telling you how typical you are for your specific place and time. National borders have been too fluid in Europe and there's been too much moving around. They're not going to be able to tell me the only other things that would interest me, such as how much "Roman", or Etruscan, or Celt, or Lombard I am, although the ones based on ancient samples can tell me pretty accurately how Anatolian Neolithic I am, or steppe herder, or WHG, so maybe with enough ancient samples someday we'll get calculators that could do this. Of course, we would have to prepare to be disappointed in that case. I'm going to be really bummed out if I have no Etruscan in me, for example, given that I've been studying them and romanticizing them since I was in university.