The figure and quotes below come from a PDF titled "Reconstructing the Human Past using Ancient and Modern Genomes". It's a summary of a thesis by Pontus Skoglund from Uppsala University.
In a follow up study (Paper IV), we obtained more sequence data from three of the previous individuals, and also sequenced 3 additional individuals associated with the TRB and 3 additional individuals associated with the PWC. Under the hypothesis that they represent different populations, we pooled the data from each group, and found that the closest modern-day groups to the TRB data and the PWC data are Sardinians, and Lithuanians, respectively. We then tested for each separate individual if they were closer to either northern Europeans (Lithuanians) or southern Europeans (Sardinians), or none. We found that all TRB-associated individuals are significantly closer to Sardinians, and all PWC-associated individuals are significantly closer to Lithuanians. In addition, we performed the same test for the 5,300 year old Tyrolean Iceman from the Italian Alps (Keller et al. 2012) and two Mesolithic Iberian individuals from La Braña, Spain (Sánchez-Quinto et al. 2012). We find that the Tyrolean Iceman is significantly closer to Sardinians whereas the Mesolithic Iberians are significantly closer to Lithuanians. This demonstrates that hunter-gatherer and farmer individuals sampled so far share genetic signatures with individuals with the same mode of subsistence from other parts of Europe, and that the Neolithic PWC individuals trace their ancestry to Mesolithic European populations.
To investigate the shared signal of the Neolithic hunter-gatherers of the PWC with Mesolithic individuals in further detail, we have also analyzed a Mesolithic individual from Stora Förvar cave on the Stora Karlsö Island in the Baltic Sea (Götherström A, Jakobsson M, Malmström H, Omrak A, Skoglund P, Storå J, unpublished results), which is very close to the locality where the PWC individuals in our studies were excavated. The individual was directly radiocarbon dated to 7,952 ± 53 BP (6,873 ± 119 cal BC).
We also found that the mtDNA sequence belonged to haplogroup U4b1, consistent with the notion that mitochondrial lineages belonging to hg U were fixed or nearly fixed in pre-Neolithic Europe (Fu et al. 2013b; Pinhasi et al. 2012). To assess whether this individual showed a similar genetic signature to other ancient individuals from the Neolithic period, we repeated the test for if the individual was closer to Northern Europeans (Lithuanians), Southern Europeans (Sardinians), or neither. We found that the Stora Förvar individual was significantly closer to Lithuanians in this analysis (Figure 7), which is consistent with a high degree of continuity from the Mesolithic in the individuals associated with the Pitted Ware culture.
All of this suggests that the East Baltic region was one of the very last refuge areas for Mesolithic hunter-gatherers in Europe, which isn't a new theory (see here).
Skoglund also has some interesting things to say about the problems in this sort of work caused by recent genetic drift and what it takes to overcome them (see pages 34-35 and 49-50). Nevertheless, I can't help but notice how closely his aDNA results correlate with my own work using modern genotype data. For instance, below are gradient maps of two ADMIXTURE components which I call Northern European and Mediterranean, respectively. The former component peaks in Lithuanians, and the latter in Sardinians. The maps come from a blog entry I posted last year titled "So who's the most European of us all?". In hindsight, that should've been "So who's the most Mesolithic of us all?".
By the way, I've also identified a third major West Eurasian component (see image here). It's modal in Georgians and Abhkasians, and I suspect it was first introduced into Europe during the Neolithic from Anatolia by farming populations associated with the Lineal pottery culture (aka. Linearbandkeramik or LBK). We'll see if that theory holds up when someone manages to sequence the genomes of LBK farmers. But it seems that Skoglund is aware that there might have been major differences between the two main types of Neolithic populations streaming into Europe.
The earliest evidence of Neolithic communities in Europe is in Cyprus and in Thessaly in mainland Greece by 8500 BP. The manner of this spread seems to indicate a reliance on maritime transportation, and as the Neolithic culture spread along the Mediterranean route into Western Europe and Iberia, a coastal route also appears to have been the main mode of movement. In contrast, the Neolithic cultures that spread into Central Europe seems to have followed the river valleys of the Danubian and Dnieper. These two paths describe the two main routes of expansion and are associated with the Impressed Ware and Cardial culture in the southwest and the Lineal Pottery complex in Central Europe, and the manner by which their spread was facilitated in terms of population history may have differed considerably (Deguilloux et al. 2012; Lacan et al. 2012; Lacan et al. 2011).
Skoglund, P. 2013. Reconstructing the Human Past using Ancient and Modern Genomes. Acta Universitatis Upsaliensis. Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Science and Technology 1069. 68 pp. Uppsala. ISBN 978-91-554-8744-7.