Our population samples included a total of 89 male subjects, currently living in Greece, who trace their grand-paternal ancestry to either the area near Phokaia (n = 31) or Smyrna (n = 58) prior to the 1923 Exchange of Lausanne. In addition 323 males living throughout Corsica who trace their paternal ancestry to the island, and 51 subjects from villages near Neolithic sites in Provence who trace their grand-paternal ancestry to Provence and the Principality of Monaco were also studied. A total of 23 of the subjects from Provence villages were from the western departments of Provence: Vaucluse and Bouche-du- Rhone, while 28 apportioned to the eastern departments: Var and Alps-de-Haute-Provence or to Monaco. Regarding the new samples introduced in this study, the Anatolian Greek component was approved by the IRB of Aristotle University, Thessaloniki, Greece. The French samples were approved by the French Committee for the Protection of Persons in Biomedical Research (CCPPRB) and the entire French collection were also declared to and approved by the French Ministry of Higher Education and Research. All subjects gave their informed consent to participate in the study. The location of the Anatolian Greek, mainland Greek, Turkish and Basques samples are shown in Figure . In addition the locations of Massalia and its trading posts and the Greek city of Alalie in Corsica are indicated. Additionally, a description of populations analyzed in this study is summarized in Table .
All 89 samples from Anatolian Greeks were genotyped using 29 Y-chromosome binary polymorphisms in a sequential manner using Y tree branching patterns to infer upstream haplogroup status. The following binary markers were genotyped: YAP, M35, V13, M78, M123, M34, M102, M9, M70, M74, M198, M269, M304, M497, M12, M241, M205, Page55, M67, M92, M530, M258, M253, M436, M223, P37.2, M423, M406, and M530. M497, Page55, Page94, and M530 are newly listed SNPs whose specifications are listed in Additional file 3
: Supplemental Table S3. Binary marker genotyping was done by RFLP assay, DHPLC or direct sequencing. Each of the Phokaia and Smyrna samples were typed at 37 YSTRs listed in Additional file 4
: Supplemental Table S4.
The 51 samples from areas near Neolithic sites in Provence had derived alleles for the following markers: V13, M34, Page94, M253, M438, M497, M530, M67, M198 and M269. In order to compare the selected 51 Neolithic samples to a larger Provence set, 368 subjects from the departments of Provence: Var (n = 68), Bouche-du-Rhone (n = 209), Vaucluse (n = 60) and Alps-de-Haute-Provence (n = 31) whose surname was blindly determined to be of French origin, were genotyped for E-V13, M406, Page94, M423, M269 and all the following J-lineages: M304, Page55, M267, M12, M410, M67, M530 and M92 as well as ten YSTRs for the J and E-V13 derived samples. DYS445 was typed in M530 derived samples (Schrack B.E., Athey T.W., Wilson J.F., 2006, The American Society of Human Genetics. Abstract). The 323 samples from Corsica were only typed for E-V13.
An AMOVA [23
] was performed using Arlequin 2000 [24
] to test the population affinities of the two Anatolian Greek samples to three mainland Greek samples (Nea Nikomedeia, Sesklo/Dimini, Lerna/Franchthi Cave), and four regions of Turkey (western Aegean, Marmara, central Anatolia and Mediterranean Turkey) [25
]. Furthermore a Multidimensional Scaling analysis (MDS) (SPSS 18.0) was performed using the Fst measure as a distance metric across the 9 populations. An AMOVA comparing the effects of geography (Asia Minor vs. Mainland Greece) and religion/language (Christian/Greek vs. Muslim/Turkish) was also calculated using these 9 populations.
To analyze the impact of the attested Greek colonization of Provence, an admixture analysis [26
] was conducted using a Basque population (n = 116) [27
] as an indigenous (non-Neolithic pre-Greek) source population and the Phokaia/Smyrna data as the Greek colonizing source represented by E-V13 frequency. As a signal of putative Neolithic immigration to Provence, central Anatolian and Mediterranean Turkey data [25
] were used. Specifically the following markers M92, M406 and J2a-(DYS445 = 6) were chosen as indicative of Neolithic ancestry. The frequencies of M92 and J2a-(DYS445 = 6) in the Basque population were estimated from their YSTR pattern [27
]. In order to assess the degree of E-V13 affinity, a 8 loci YSTR network using Phokaia, Smyrna, Provence and Corsica samples was constructed [28
]. Networks were constructed by the median joining method using Network 188.8.131.52, where ε = 0 and microsatellite loci were weighted proportionally to the inverse of the repeat variance observed in each haplogroup[29
]. Coalescent times for E-V13 based on the following 8 loci DYS19, DYS389I, DYS389II, DYS390, DYS391, DYS392, DYS393 and DYS439 were computed using the methodology of Zhivotovsky et al.[30
] as modified according to Sengupta et al.[31
]. A microsatellite evolutionary effective mutation rate of 6.9 × 10-4
per 25 years was used [30