I have made two PCA plots using new Baltic samples published by Mittnik et al. 2018, but as far as I can see with a different approach. On my plots modern populations are placed at first hand on grounds of ancient samples, giving, in my humble opinion, a historically correct approach. I have also standardized sample sizes of modern populations to minimize the effects of modern genetic drift, which of course, if exists, will lead to comparison error due to irrelevant PCA-components. Thirdly I have here two regional view; Eurasian and European. The Asian effect on the Eurasian picture is pronounced in certain modern populations implying Asian admixture events after the Bronze Age Europe.
Close-up of Eurasian picture:
On the Eurasian picture the first dimension (X) represents East-Asian and Siberian maximums and the second one (Y) represents ancient farmer maximum. So the picture isn't relevant for checking dimensions inside present-day Europe. It is easy to understand that the flat left represents Asian/Siberian minimum regardless of later European structures. Similarly the the vertical axis represents variation between ancient farmers and hunter-gatherers rather than structures inside present-day Europe. The third picture (Europeans only) represents historically correct variation between ancient farmers and hunter-gatherers and exclusively to these variations ancient Steppe populations.