Light Version
Dark Version
The graphs below show genetic modellings of Greeks, Iranians, Kurds, Anatolian Turks, Armenians, Serbians and Bulgarians so we can get a better grasp of the genetic structure of the West Asia and Balkans. Ethogenetic formations of the given populations were taking into consideration while running the models. We used G25 PCA coordinates and Fast Monte Marlo simulation to get the results. Samples and fit values for each model can be found in this spreadsheet.
There has been a long lasted discussions around the genetic structure of Anatolian Turks, more precisely on their genetic affinity with other Turkic speaking peoples in Central Asia and Siberia. In order to clarify this stiuation we need to start asking the right questions first because most people (including the academicans) have a misconception in their minds about what Turkic peoples are (and were) from the both historical and genetic point of view. First of all, comparing Anatolian Turks with contemporary Kazakh, Kyrgyz, Altai and etc. people is factually wrong and it leads people into false conclusions about their genetic relation with Anatolian Turks. This simply because of the centuries long separate history and ethnogenetic formations between these various Turkic peoples. After taking these facts into account the right question should be: “How much Oghuz Turkish genetic heritage Anatolian Turks have?” since it was the Oghuz Turks who started to migrate into Anatolia in 11th century and lay the foundation of Turks in Turkey today.
There are only few worth mentioning genetic studies done on Turkish people. All of them are old and somewhat outdated considering the rapid development in population genetics and ancient DNA field in the last few years. The first one is from 2004. A Y-DNA study called Excavating Y-chromosome haplotype strata in Anatolia led by Cengiz Cinnioglu. In this study the samples are randomly collected from blood banks without ethnic separation and on top of that there was a huge inequality between the sampled regions in terms of sample sizes, which were not in favor of ethnic Turkish demographics. Eastern Anatolia, Eastern Black Sea region and a metropole like Istanbul has a much bigger share in the total sample size. The study also suggested that only East Eurasian related Y-DNA Haplogroups Q, O and C (they did not even include N/N1 because it was thought be an European related haplogroup back in 2004) are an indicator of Central Asian/Turkic ancestry which is already proven wrong by the recent ancient DNA studies. In comparison, the samples we have gathered from people whose known paternal and maternal lineages are ethnic Turks are shown here. As we can see the results are vastly different.
The second famous study was conducted by Ugur Hodoglugil in 2012. An autosomal DNA study called Turkish Population Structure and Genetic Ancestry Reveal Relatedness among Eurasian Populations. The study used Turkish samples collected from Aydın, Adana, İstanbul, Balıkesir, Kayseri and Trabzon provinces for a heart health study in 1995 by Mahley et. al. Mahley and later Hodoglugil did not make any ethnic separation between these sampled individuals either, therefore the samples included Kurdish, Circassian, Arab and Balkan immigrant minorities as well. Alongside these mixed samples, modern Kyrgyz people used as a source population in the study for estimating the Central Asian admixture among Turkish people. They found out that “Turks” have 15% Central Asian-like genetic heritage. Which is not that surprising considering the samples and method they used.
An another autosomal DNA study published in 2014 by Can Alkan. This time they did a WGS (Whole Genome Sequencing) on 16 individuals from various regions of Turkey “irrespective of their mother-tongue/ethnicity”. According to the analysis in the study the weigh of East Asian (?) ancestry was estimated around 21% among the sampled individuals. Though they also used modern populations in the mentioned analysis, they noted that genetic profile of the Turks who migrated into Anatolia in 11th century is still unknown hence the estimations are based on speculations.
Granted, at the time these studies were published we had not many ancient DNA samples. Also the primary aim of these studies seems to be all peoples of Turkey rather than ethnic Turks. Still, the way they presented their conclusions about Central Asian genetic heritage among “Turks” was heavily misleading.
As mentioned earlier, if we are going to estimate Central Asian genetic heritage among Anatolian Turks, the healthiest way to do it is by using Oghuz Turkish samples from Seljuq period (11th-13th centuries). Eventhough we don’t have any Oghuz DNA samples yet, we do have samples from other medieval Turkic peoples such Gokturks, Karakhanids, Karluks, Kipchaks and Kimaks and they are currently the best proxies we have for representing Oghuz Turkish genetic heritage.
By using Global25 PCA coordinates (25 eigenvectors scaled) and Fast Monte Carlo simulation methods we ran several models with the regional averages of Anatolian and Balkan Turks (most of them provided by our project) and Medieval Turkic populations. We used Central Anatolian Greeks, Trabzon Greeks, Greeks from Kos Island, Armenians, Georgians, Copper Age Iran (Hajji Firuz), Bulgarians and Macedonians to represent pre-Turkic Anatolia and Balkans. An interactive map that shows the results and model details can be found here.
The graph below summarizes the results of Anatolian Turks:
As we can see here, when the correct and representative samples are used the results can change significantly. We will get more accurate results once there will be proper Oghuz Turkish and Byzantine Era Anatolian Greek/Armenian samples but so far the data we have give us a near-solid results.
See also:
