Challenging the Lack of Thracian Ancestry in Contemporary Bulgarians: A Reanalysis of Svetoslav Stamov’s f3 Approach with Additional Populations.

Abstract

This study aims to re-examine the conclusion that contemporary Bulgarians do not have significant Thracian ancestry, as stated in a 2021 preprint, by utilizing the same f3 approach but including additional populations, specifically Russia_Alan.SG, Bulgaria_Medieval, Bulgaria_EIA and Macedonia_IA. The aim is to test whether the presence of these additional populations alters the conclusions of the original study regarding the absence of Thracian ancestry in contemporary Bulgarians. The same software used by the previous study, ADMIXTOOLS version 2 and the updated version V54.1 of the „Allen Ancient DNA Resource (AADR)“ dataset, is also used in this study. The results of the f3 analysis indicate that the Russia_Alan.SG and Bulgaria_Medieval populations exhibit closer genetic affinities to Icelandic and Greek populations respectively than to contemporary Bulgarians. Furthermore, the contemporary Bulgarian population is significantly more closely related to the Bulgaria_EIA and Macedonia_IA, representing ancient Balkan DNA, than to the Bulgaria_Medieval and Russia_Alan.SG populations, purporting to represent proto-Bulgarian DNA.

Introduction

In a 2021 preprint called „Significant (Z|-4|) admixture signal with a source from ancient Wusun observed in contemporary Bulgarians“, Svetoslav Stamov presents an analysis of ancient DNA data with the aim of providing insights into the history of the Bulgarians. One of the conclusions presented in the study is that „contemporary Bulgarians do not harbor Thracian-specific ancestry, since ancient Thracian samples share more SNPs with contemporary Greeks and even contemporary Icelanders than with contemporary Bulgarians“. However, the current study aims to re-examine this conclusion by utilizing the same f3 approach as Stamov, but including additional populations, specifically Russia_Alan.SG and Bulgaria_Medieval. We also utilize the f2 approach. We test whether the different populations alter the conclusions drawn in the original study regarding the absence of Thracian ancestry in contemporary Bulgarians.

Dataset and software for the analysis

It is a common practice in Svetoslav Stamov’s work to make conclusions based on limited sample size. In this instance, his assertion that there is „no substantial contribution coming from ancient Thracians to contemporary Bulgarians“ is based on a single DNA sample from the Iron Age in Bulgaria.

In order to evaluate the validity of Stamov’s conclusions, the same modern populations from the „Simons Genome Diversity Project“ will be utilized in this analysis. This will allow for a direct comparison to the data utilized by Stamov and an assessment of the accuracy of his conclusions. In order to maintain brevity and clarity in the analysis, the following modern populations will be included in the analysis: Dinka, Bulgarian, Yakut, Icelandic, Greek, and Polish. We are utilizing the same dataset as Svetoslav Stamov, in the updated version V54.1 from November 16th, 2022; and the same software, ADMIXTOOLS version 2.

F3 analysis

In order to further examine the relationship between the study populations, the f3 statistic was applied using the methodology proposed by Patterson et al. in „Ancient Admixture in Human History“, 2012. We included the Russia_Alan.SG and Bulgaria_Medieval populations as additional reference groups. The results of this analysis indicate that the Russia_Alan.SG and Bulgaria_Medieval populations exhibit closer genetic affinities to Icelandic and Greek populations, respectively, than to contemporary Bulgarians. For f3, the higher the f3 score, the greater the genetic similarity between the second and the third population.

These are the results:

   pop1     pop2              pop3           est      se     z     p      n
   <chr>    <chr>             <chr>        <dbl>   <dbl> <dbl> <dbl>  <dbl>
 1 Dinka.DG Bulgaria_IA       Yakut.DG     0.127 0.00193  65.8     0 385997
 2 Dinka.DG Bulgaria_IA       Icelandic.DG 0.161 0.00221  72.6     0 381503
 3 Dinka.DG Bulgaria_IA       Bulgarian.DG 0.159 0.00218  73.1     0 381462
 4 Dinka.DG Bulgaria_IA       Greek_1.DG   0.162 0.00246  66.0     0 380652
 5 Dinka.DG Bulgaria_IA       Polish.DG    0.159 0.00248  64.3     0 380895
 6 Dinka.DG Russia_Alan.SG    Yakut.DG     0.131 0.00161  81.7     0 589601
 7 Dinka.DG Russia_Alan.SG    Icelandic.DG 0.157 0.00177  88.9     0 581891
 8 Dinka.DG Russia_Alan.SG    Bulgarian.DG 0.154 0.00169  91.4     0 581822
 9 Dinka.DG Russia_Alan.SG    Greek_1.DG   0.156 0.00181  86.3     0 580489
10 Dinka.DG Russia_Alan.SG    Polish.DG    0.154 0.00185  83.0     0 580937
11 Dinka.DG Bulgaria_Medieval Yakut.DG     0.132 0.00174  75.8     0 492705
12 Dinka.DG Bulgaria_Medieval Icelandic.DG 0.163 0.00198  82.4     0 486820
13 Dinka.DG Bulgaria_Medieval Bulgarian.DG 0.159 0.00184  86.6     0 486761
14 Dinka.DG Bulgaria_Medieval Greek_1.DG   0.161 0.00209  77.0     0 485694
15 Dinka.DG Bulgaria_Medieval Polish.DG    0.159 0.00214  74.3     0 485999

Our analysis of the genetic data utilizing the software ADMIXTOOLS 2 reveals a proximity of the Russia_Alan.SG and Bulgaria_Medieval populations, which are thought to represent the ancestral proto-Bulgarians, to the Icelandic and Greek populations when compared to contemporary Bulgarians. This observation is consistent with the genetic proximity observed between the Bulgaria_IA population and the aforementioned populations. Based on this, if we apply the logic used by Stamov in his study, it can be inferred that contemporary Bulgarians do not exhibit genetic affinity specific to the proto-Bulgarian population.

F2 analysis

An alternative approach to infer genetic relatedness between populations is the use of f2 statistics, which do not rely on the inclusion of an outgroup population. Specifically, the lower the f2 score, the greater the genetic similarity between populations. Here we add the populations Bulgaria_EIA and Macedonia_IA, published in 2022.

These are the results:

   pop1              pop2             est       se
 1 Bulgaria_EIA      Bulgarian.DG 0.00316 0.000443
 2 Bulgaria_EIA      Greek_1.DG   0.00353 0.000882
 3 Bulgaria_EIA      Icelandic.DG 0.00582 0.000492
 4 Bulgaria_EIA      Polish.DG    0.00689 0.000902
 5 Bulgaria_EIA      Yakut.DG     0.0375  0.000436
 6 Bulgaria_IA       Bulgarian.DG 0.157   0.000945
 7 Bulgaria_IA       Greek_1.DG   0.156   0.00129 
 8 Bulgaria_IA       Icelandic.DG 0.159   0.000946
 9 Bulgaria_IA       Polish.DG    0.160   0.00136 
10 Bulgaria_IA       Yakut.DG     0.191   0.000810
11 Bulgaria_Medieval Bulgarian.DG 0.0278  0.000722
12 Bulgaria_Medieval Greek_1.DG   0.0273  0.00113 
13 Bulgaria_Medieval Icelandic.DG 0.0278  0.000794
14 Bulgaria_Medieval Polish.DG    0.0295  0.00120 
15 Bulgaria_Medieval Yakut.DG     0.0584  0.000746
16 Macedonia_IA      Bulgarian.DG 0.00220 0.000408
17 Macedonia_IA      Greek_1.DG   0.00195 0.000863
18 Macedonia_IA      Icelandic.DG 0.00397 0.000477
19 Macedonia_IA      Polish.DG    0.00523 0.000896
20 Macedonia_IA      Yakut.DG     0.0355  0.000405
21 Russia_Alan.SG    Bulgarian.DG 0.0138  0.000570
22 Russia_Alan.SG    Greek_1.DG   0.0136  0.000981
23 Russia_Alan.SG    Icelandic.DG 0.0149  0.000619
24 Russia_Alan.SG    Polish.DG    0.0170  0.00105 
25 Russia_Alan.SG    Yakut.DG     0.0418  0.000556

The results of the pairwise f2 statistics indicate that:

• The population of Bulgaria_EIA is more closely related to the contemporary Bulgarian population than to the contemporary Greek population.
• The population of Bulgaria_EIA, Bulgaria_IA, and Macedonia_IA exhibit a closer relationship with the contemporary Bulgarian and Greek populations, whereas the Icelandic, Belarusian, and Yakut populations exhibit a more distant relationship.
• The contemporary Bulgarian population is significantly more closely related to the Bulgaria_EIA and Macedonia_IA (representative of ancient Balkan DNA) than to the Bulgaria_Medieval and Russia_Alan.SG populations (purportedly representative of proto-Bulgarian DNA)
• The Bulgaria_Medieval and Russia_Alan.SG populations (purportedly representative of proto-Bulgarian DNA) exhibit a closer relationship with the contemporary Greek population than with the contemporary Bulgarian population.

Conclusion

The study by Svetoslav Stamov, published in a bioRxiv preprint in 2021, makes incorrect claims about the genetic makeup of contemporary Bulgarians. Using a newer version of the same dataset and similar analytical methods, this critique demonstrates that contemporary Bulgarians do in fact harbor significant ancestry from ancient Thracians. Furthermore, the analysis provides evidence that Bulgaria_Medieval and Russia_Alan.SG, the populations cited by Stamov as being of proto-Bulgarian origin, are not more closely related to contemporary Bulgarians than to other populations such as Greeks and Icelanders. These findings refute Stamov’s conclusion that contemporary Bulgarians do not harbor Thracian-specific ancestry. This critique serves as a reminder that conclusions made from genetic data should be made with care and scrutiny, especially when based on small sample sizes or limited genetic markers.