This PLoS paper is the first academic study to use high density genome-wide SNPs to analyze samples from the Baltic States. It shows some interesting and perhaps surprising outcomes, including the following:
- Estonians are closer to Poles, Latvians and Lithuanians, in that order, than to Finns.At least that's my interpretation of the various analyses featured in the paper and supplementary data. Admittedly, most of these aren't results I would've bet on before seeing this study, but now that I have, they do make a lot of sense.
- Finns are distant from all Europeans, including fellow Northeastern Europeans, but closest to Estonians, Poles and Swedes.
- Russians from Tver are more southern and western genetically than Latvians and Lithuanians, despite being located east and northeast of them, respectively.
- Poles from West Pomerania (presumably largely of eastern Polish descent due to post-WWII re-settlements) are closer to Estonians and Austrians than to Lithuanians and Latvians.
- The most northern (or least southern) samples in terms of genetic affinities are Latvians, Lithuanians and Finns.
For instance, the large genetic distances between Finns and other Europeans, including even their immediate neighbors and fellow Finnic-speakers Estonians, can be explained by strong founder effect and resulting genetic drift across Finland. However, some Finns do overlap with Estonians and Swedes on the aforementioned PCA and MDS plots, which probably means these individuals are less drifted and/or carry recent admixture from outside of Finland.
The fairly high affinity between Estonians and Poles might be due to shared deep ancestry from near the Baltic Sea as well as similar levels of recent Western European influence. I suspect it's the western admixture which makes Estonians and Poles look more similar to each other in the inflation factor lambda table than to their Baltic neighbors Latvians and Lithuanians, respectively, who probably carry lower levels of this admixture.
It's also possible that Latvians and Lithuanians are being pushed east because of ancient influence from their Eastern Baltic-speaking ancestors, who lived in what is now Russia during the Iron Age. On the other hand, the Russians from Tver are Slavic-speakers whose ancestors arrived in Northwestern Russia from the Slavic homeland somewhere in East Central Europe during the early Middle Ages. These factors might explain the more southern and western genetic character of these Russians compared to the Balts, which is obviously an outcome that doesn't fit geography.
The results for the Swedes are also out of whack with geography. On the PCA/MDS plots they basically look like Central Europeans with varying degrees of Finnish admixture. However, all other Germanic-speaking samples come out very Central European too, so perhaps this says something about the origins of this linguistic group? The Finnish admixture in Sweden is easily explained by heavy migration of Finns to Scandinavia during historic times, although ancient links between Finland and Sweden might also be partly responsible.
Nelis et al. also investigate genetic variation within several European countries. They show that PCA can separate samples from northern and southern Italy with almost 100% accuracy. Germany and Estonia also show some differentiation along the north-south axis, but not to the same extent. On the other hand, samples from northwest and southeast Czech Republic are almost inseparable, and also very similar to Austrians.
Nelis M, Esko T, Mägi R, Zimprich F, Zimprich A, et al. (2009) Genetic Structure of Europeans: A View from the North–East. PLoS ONE 4(5): e5472. doi:10.1371/journal.pone.0005472
Northwestern (or rather Eastern) Poles compared to other Europeans using more than 270K SNPs