Koban culture genome-wide and archeological data open the bridge between Bronze and Iron Ages in the North Caucasus

Abstract

The North Caucasus played a key role during the ancient colonization of Eurasia and the formation of its cultural and genetic ancestry. Previous archeogenetic studies described a relative genetic and cultural continuity of ancient Caucasus societies, since the Eneolithic period. The Koban culture, which formed in the Late Bronze Age on the North Caucasian highlands, is considered as a cultural “bridge” between the ancient and modern autochthonous peoples of the Caucasus. Here, we discuss the place of this archeological culture and its representatives in the genetic orbit of Caucasian cultures using genome-wide SNP data from five individuals of the Koban culture and one individual of the early Alanic culture as well as previously published genomic data of ancient and modern North Caucasus individuals. Ancient DNA analysis shows that an ancient individual from Klin-Yar III, who was previously described as male, was in fact a female. Additional studies on well-preserved ancient human specimens are necessary to determine the level of local mobility and kinship between individuals in ancient societies of North Caucasus. Further studies with a larger sample size will allow us gain a deeper understanding of this topic.

Introduction

The Caucasus, the region between the Black and Caspian Seas, is considered as an important bridge for the migration of ancient people to Europe and Asia, since the Upper Paleolithic [1]. Moreover, unique climatic and geographical conditions made this region suitable for the origin and development of ancient archeological cultures [2]. Earlier studies showed genetic continuity between the culture bearers that inhabited the Caucasus during the Bronze and Iron Ages [3, 4]. In addition, despite external influences, many local ancient societies have maintained cultural continuity over the centuries [5, 6].

The Koban culture of the Late Bronze and Early Iron Ages (the end of the 2nd–the middle of the 1st millennium BC) is considered as a “bridge” between the Bronze Age cultures and modern ethnic groups in the North Caucasus [7]. This archeological culture acquired its name from the village of Verkhny Koban (in modern North Ossetia, Russia), near to which the first bronze artifacts from this culture were discovered in the middle of the 19th century. Other archeological sites, such as cemeteries and settlements, of the Koban culture were found on both sides of the Great Caucasus Range (republics of Ingushetia, Kabardino-Balkaria, Karachaevo-Cherkessia, North Ossetia–Alania and Chechnya, and Stavropol’ Krai in Russia and modern South Ossetia and Georgia) (Fig. 1) [6–10].

Fig. 1. Geographical map of the Caucasus showing the archeological sites (red circles—Koban culture: Zayukovo-3 and Klin-Yar III burial sites; blue circle—early stage of Alanic culture: “Bratskiye 1-ye Kurgany” barrow cemetery) where human samples were collected for the ancient DNA analysis.

Koban culture distribution in the Caucasus between 1100 and 400 BC marked by yellow area.

During its existence, the Koban culture bearers closely interacted with the states and peoples of Transcaucasia and the Near East and with Cimmerians and later Scythians, whose cultural influence can be traced in Koban culture [7]. The Koban societies knew advanced metallurgy (Fig. 2A;  2B) as well as developed terraced agriculture [11] and cattle husbandry. The Koban culture is considered by many researchers as an autochthonous archeological culture of the North Caucasus. It is believed that the Koban culture bearers played an important role in the formation of the genetic pool of modern North Caucasian ethnic groups [6, 8, 12].

Fig. 2. Koban culture.

A Zayukovo-3 burial ground, burial 72. Scabbard chape, 5th century BC. Photo by A.A. Kadieva. B Zayukovo-3 cemetery, burial 33. Photo by A.A. Kadieva.

Archeogenetic studies of the Koban culture bearers have not been carried out until recently [13, 14]. Mitochondrial genome variability in Koban culture individuals demonstrates the genetic continuity of this marker among the ancient communities and modern ethnic groups of the North Caucasus and corresponds with previous studies [1, 3, 4]. Indo-European mitochondrial haplogroups, which have been widespread in Western Europe since the Neolithic, have also been dominant among the studied Koban culture individuals [14]. At the same time, the Y-chromosome markers in the Koban culture bearers are more heterogeneous and apparently more dependent on the male founder effect that was previously described in the Caucasus [1, 15]. The G2a1a and R1b Y-chromosome haplogroups that were reported in Koban culture individuals [13] are widely present across individuals from ancient Eurasian archeological cultures, including Caucasian (Yamnaya, North Caucasian, and Catacomb cultures) [3, 16] as well as modern Ossetians, Balkars, and Karachays [15, 17].

At the turn of the Late Bronze and Iron Ages, Western Sarmatians and Western Scythians, who had low genetic distances between each other [18], made a significant cultural contribution to North Caucasus ancient societies. In particular, such influence can be traced in the Koban and subsequent archeological cultures [7]; in addition, archeological findings suggest a gradual shift of the Koban culture from the west to the east part of the Caucasus over time [19].

In this study, we use deep genome sequencing of ancient DNA because previous data suggest both the genetic continuity of the North Caucasus and a genetic exchange between the Koban culture bearers and individuals who inhabited the adjacent steppe during the same historical period [13, 14]. Both genome-wide data and archeological findings clearly show that the Koban culture bearers have continuity with other ancient (Alans) and modern ethnic groups of the Caucasus. This study sheds light on the genetic and cultural transition between the Bronze and Iron Ages in the North Caucasus and on the formation of genetic diversity of the modern local ethnic groups.

Results

The total number of reads generated for five Koban culture and one Alanic culture individuals ranged from 123,479,842 to 453,064,190 per DNA library. Despite using ancient DNA facilities, 44.4% – 48.3% of reads were filtered (as contamination) from the sequenced datasets. The number of filtered reads ranged from 63,779,580 to 251,521,901 for the DNA libraries analyzed. PALEOMIX filtered reads that contained postmortem cytosine deamination patterns specific to ancient DNA (Fig. S1–S6; Supplementary Table 5) were mapped to the human reference genome (GRCh37). The average genomic coverage per analyzed individual varied from 0.034× to 0.292×. The number of endogenous reads ranged from 0.29% to 10.08%. The mapping statistics is shown in Supplementary Table 2. Mitochondrial DNA and X-chromosome contamination tests showed that the generated genomic datasets were suitable for downstream analyses (the number of nuclear genome SNPs varied from 28,680 to 243,300 for the studied individuals) (Supplementary Tables 3–5). The genomic datasets for Zayukovo-3:ID79/1 and Klin-Yar III:ID 355 have significant levels of contamination, as indicated in Supplementary Table 3–4. Despite this, these specimens were not excluded from the analysis, but any conclusions drawn from them must be viewed with caution. Moreover, Zayukovo-3:ID79/1 male individual is marked as contaminated in PCA and ADMIXTURE analyses (see below), based on hapCon tool [20] results. In addition, the ancient DNA analysis led to the revision of the previous anthropological sex determination [21] of Klin−Yar III:ID355 individual to female (Supplementary Table 4).

The genomic data for previously published ancient Caucasus and Western Eurasian as well as modern Eurasian populations were obtained from the Allen Ancient DNA Resource (AADR) database (https://reich.hms.harvard.edu/allen-ancient-dna-resource-aadr-downloadable-genotypes-present-day-and-ancient-dna-data). We merged the SNPs dataset generated in our study with selected SNP datasets from the AADR database (Supplementary Dataset 1). This combined dataset was used for principal component analysis (PCA), ADMIXTURE, and f3- and f4-statistics analyses to assess the genetic relations between Koban culture bearers, other ancient Caucasus individuals, ancient individuals from the surrounding steppe, and the modern ethnic groups inhabiting this region.

Based on the PCA plot, we clearly showed that the ancient Koban culture bearers are in close genetic proximity to the ancient individuals from the Kura-Araxes and Maykop Bronze Age cultures on one side and the Alanic culture on the other side (Fig. 3; Fig. S7). At the same time, the single Alanic culture individual (“Bratskiye 1-ye Kurgany” barrow cemetery:ID1402) from our experiment is wedged between the Scythian and Alanic culture bearer genomes that were previously published by Damgaard et al. [4] (Fig. 3; Fig. S7). It is interesting to note that all four Koban representatives from Zayukovo-3 are grouped with Alanic culture individuals, while it seems that the individual from Klin-Yar III:ID355 has a place close to Western Scythians (Fig. 3; Fig. S7) [22]. This latter individual (Klin−Yar III:ID355) was, moreover, buried in a way that was unusual for the western variant of the Koban culture. Despite anthropological sexing as male, their grave contained female-associated types of artifacts, and the low zinc concentration in their bones suggests specific status of the mineral bone part (probably a largely vegetarian diet), in contrast to other Koban individuals buried at Klin-Yar III. Despite the high contamination level, genomic data suggested that Klin-Yar III:ID355 individual was female. Koban culture representatives have low genetic distances from the present-day, isolated Hamsheni Armenian ethnic group, among the modern populations on the PCA plot (Fig. S8) [23]. We also show that Caucasian Bronze and Iron Ages culture bearers (including Koban culture) play an important role in forming the gene pool of present-day West Eurasians or have the same genetic ancestry as them (Fig. S9).

Fig. 3. Principal component analysis (PCA) of Koban culture individuals.

Genetic structure of Koban culture bearers (green squares highlighted by green sphere) compared to those of other archeological cultures of the North Caucasus: Alans (Russia)—navy blue rhombs and grey square highlighted by blue sphere; Cimmerian (Moldova and Ukraine)—blue circles and orange triangles; Sarmatians (Russia)—red triangles, red rhombuses, orange squares, and orange rhombuses; Scythians (Moldova and Ukraine)—blue squares, yellow triangles, and yellow squares.

The previously published archeological and Y-chromosome data suggest that the Scythian invasions had a significant influence on the cultural as well as the genetic legacy of the Koban culture bearers [7, 8, 13]. ADMIXTURE analysis, which was carried out for the Koban individuals, showed that they share a similar genetic profile with Alans, and, compared to the individuals of the Caucasian Bronze Age (Maykop and Kura-Araxes) cultures from the highlands, they had some genetic influence from steppe nomads (Scythians for the Koban population and Sarmatians for the Alanic population) (Fig. 4; Fig. S10).

Fig. 4. ADMIXTURE profiles (k =  12) of ancient humans inhabiting the Caucasus and adjacent regions during the Paleolithic and Neolithic periods, Bronze and Iron Ages (Kura-Araxes, Maykop, Koban, and Alanic culture bearers from foothills, Yamnaya culture bearers, Scythians, and Sarmatians).

Ancient data projected onto the modern population genetic data.

The f3-statistics with the ancient populations of the Caucasus and adjacent regions as the target resulted in a significantly negative f3 (|Z score | > 3) when the combined Koban population SNPs dataset was used as one potential source and the Alanic historical individuals as the second potential source. An f3-statistics analysis showed the genetic influence from the Western Eurasian nomads (possibly, ancestors of the Sarmatians) and Koban culture bearers on the gene pool of ancient Alanic culture individuals (Fig. 5A). Moreover, an f4-statistics analysis demonstrated the gene flow from Koban to Alanic individuals (Fig. 5B). The f3-statistics analysis, which is based on allele frequency correlations between the ancient and modern human populations of the Caucasus and adjacent geographical regions, showed that (compared to other modern Caucasus ethnic groups) the modern Kumyk and Lezgin populations have a higher genetic legacy from the ancient Koban culture and Sarmatian individuals who were analyzed in this study (Fig. S11).

Fig. 5. f-statistics for Koban and Alanic culture bearers.

A F3-statistics for Koban and Alanic populations. The genetic component associated with Sarmatians and Koban culture bearers is also possessed in Alanic culture individuals. F3 (Koban; Source2; Russia_Alan). The statistically significant genetic influence from the Koban and Alanic individuals (Z score < −3) is marked in blue. Target populations are on the Y-axis. B Results of f4-statistics. F4 (Koban; Test; Russian_Alan; Yoruba). Significant Z-scores are highlighted in blue. Target populations are on the Y-axis.

The previously obtained data on the mitochondrial and Y-chromosome haplogroups as well as the genomic SNPs data of people who inhabited the North Caucasus at the turn of the Bronze and Iron Ages indicate cultural and genetic continuity over the centuries [3, 13]. At the same time, there has been, since the time of the Koban culture individuals, an increase in the steppe genetic component in the representatives of the North Caucasus mountains, which was previously (during the Bronze Age) observed only in the Caucasian foothills (Fig. 4; Fig. S10). Despite a small sampling size, we carried out a kinship analysis between the Koban culture bearers from the Klin-Yar III and Zayukovo-3 burials. Unfortunately, we did not find a kinship between non-contaminated genomic datasets of Koban culture individuals (Fig. S12).

Discussion

The Koban archeological culture is one of the brightest phenomena of the Late Bronze Age–Early Iron Age in the Caucasus. However, the origin of the Koban culture is still unresolved, even though archeological studies of the Koban culture began in the second half of the 19th century [6, 7, 10].

Koban societies established high-level agriculture in wide areas of the Kislovodsk Basin, including artificial terracing on the foothill slopes [11]. These extensive methods of agriculture, together with the global cooling (the Homeric Minimum) and humidification of the climate, which occurred approximately from 800 to 600 BC [24], led to disastrous consequences for this archeological culture in the Kislovodsk microregion. In this historical period, Koban artifacts disappear from the Kislovodsk Basin but appear in neighboring territories. The archeological sites of the Koban culture of the foothills and the mountain zone of Kabardino-Balkaria have more recent dates. One of these archeological sites is the Zayukovo-3 cemetery, where 87 burials of the Koban culture (7th-4th centuries BC) have been studied to date. Migration of Koban culture individuals to the east can be traced both by certain types of artifacts and by features of the funerary rite, primarily the crouched deposition of the bodies in graves, which was practiced by the Koban culture in the Kislovodsk Basin and was possibly transferred to the Baksan Gorge and other regions of Kabardino-Balkaria [7]. Cultural and population links between those regions are confirmed by an analysis of the isotope signals of the paleodiet of individuals buried in the Klin-Yar III and Kichmalka II cemeteries, the latter situated in between the former site and Zayukovo-3 [25]. Thus, it is necessary to carry out kinship analyses of ancient people among and inside certain Koban culture archaeological sites on both sides of the Great Caucasus Range in further. These analyses can shed light on the social structure and relationships within these societies, providing valuable insights into their origin and genetic connections.

It was traditionally suggested that the Koban culture was descended from the Middle Bronze Age North Caucasian culture [12]. Subsequently, radiocarbon dates showed that these two had been separated in time by almost a thousand years [7, 9, 26]. Another possible link between the North Caucasian and Koban cultures was thought to be the “post-Catacomb” burials that had been identified in the foothills of the North Caucasus [27–29]; however, even in this case, there is a gap of about 500 years [10]. An additional suggestion was that the Koban culture initially emerged in the territory of modern South Ossetia and subsequently spread to the northern side of the Caucasus Mountain range. A significant number of burial complexes have been studied there, demonstrating a continuous development of local cultures from the 17th/16th century to the 7th century BC [9, 30]. Nevertheless, few archeologists believe in direct links of the local cultural center in South Ossetia to the North Caucasus in the Late Bronze Age [10]. It seems that the Koban culture had a polycentric origin with different cultural roots from the various Transcaucasian and adjacent regions [6, 7].

To date, only a few studies on the genetic ancestry of the Koban culture have been published. Diverse mitochondrial haplogroups—I1, J1c, H1e, H20a, HV1, N, R6, T1a, W5a—were reported for 11 human individuals from the Klin-Yar III and Zayukovo-3 archeological sites [13]. Two of them also had the Y-chromosome R1b haplogroup, which is the same haplogroup that has been observed widely across Eurasia [16], including in ancient Cimmerian, Scythian, and Sarmatian nomads [22]. The Y-chromosome G2a haplogroup, which is noted in Near Eastern and European Neolithic cultures [31, 32] and is common in modern Caucasian ethnic groups [15, 17], was also reported among the analyzed Koban individuals. Another study, which was focused on the variability of hypervariable region 1 of mtDNA of 71 ancient human individuals (Koban, Middle Sarmatian and Alanic cultural bearers as well as North Caucasian population of the Sarmatian period), described a high mitochondrial diversity in this region with a predominance of West Eurasian mitochondrial haplogroups [14].

Our study presents the first genome-wide analysis of five prehistoric Koban culture individuals who had been buried in the Zayukovo-3 and Klin-Yar III cemeteries (previously described on the mtDNA and Y-chromosome levels in Boulygina and colleagues [13]) and one early Alanic culture individual from the “Bratskiye 1-ye Kurgany” barrow necropolis. The autosomal DNA profiles of the five Koban individuals show their close genetic connections with the individuals from the Caucasian Bronze Age archeological cultures (the Kura-Araxes and Maykop cultures) on the one hand and Iron Age cultures (the Alanic culture) on the other hand. This result shows the role of the Koban culture bearers as an ancient genetic “bridge” between the Bronze and Iron Ages in the North Caucasus (Fig. 3; Fig. S7–S9).

In addition, the admixture profiles of the Koban culture individuals demonstrate that steppe nomads (Scythians) influenced the Caucasus populations not only culturally but also genetically. Moreover, the Klin-Yar III individual whose grave had several features that were unusual for the Koban culture seems to have a close cultural and genetic connection with the Black Sea Scythians (Fig. 3).

The genetic profile of the analyzed Koban culture and Alanic individuals are mostly the same in terms of the ratio of genetic contributions (Fig. 4; Fig S10). They mainly consist of the genetic profile of the local Caucasus population dating back to the Bronze Age, with a significant genetic contribution from representatives of the steppe nomad populations (e.g., Scythians and Sarmatians). This result fully corresponds to the modern archeological view on the origin of the North Caucasus Alans [14, 33].

The accurate determination of the sex of skeletal remains is widely recognized as crucial in the field of archaeology. This determination serves as a prerequisite for gathering additional information, such as the individual’s age and height [34]. Previously, Klin-Yar III:ID 355 individual was classified as male based on anthropological evidence, despite the limited presence of animal protein in their diet [35] and the discovery of female-associated grave goods [21], including a bronze pin, glass beads, and bronze tubes, within the burial site. Here, through the implementation of genome-wide data, we revise this assumption and demonstrate that Klin-Yar III:ID 355 was, in fact, female.

A comparative genomic analysis of the Koban individuals and modern-day human populations shows that the Koban culture bearers left their genetic ancestry in the modern Kumyk and Lezgin populations (Fig. S11) and probably had the same genetic roots as the isolated Hamsheni Armenian ethnic group (Fig. S8).

Conclusion

The genomic data obtained from the Koban culture bearers, together with previously published ancient human genome datasets of the North Caucasus [3, 4], demonstrate the genetic continuity of ancient populations in this region where archeological cultures succeeded one another, with cultural and genetic invasions from the adjacent steppe. We suggest that future genomic studies, together with other modern methods of the natural sciences available for the investigation of archeological sites, will shed light on the genetic diversity of the populations behind these ancient archeological cultures, on the degree of kinship between them, and on previously unknown historical and population events in the North Caucasus.

Material and methods

Koban Culture description: Zayukovo-3 and Klin-Yar III burial sites

Klin-Yar III is an archeological site (ancient cemetery) located near the town of Kislovodsk (Stavropol Krai, Russia) in the foothills of the North Caucasus. Hundreds of ancient graves have been excavated here after its discovery, most of them belonging to the (i) Koban culture. However, some of them belong to the (ii) Late Iron Age Sarmatian period as well as the (iii) early medieval Alanic culture. Archeological fieldwork on this site (cemetery area III) in 1994 – 1996 recorded graves from all three periods. More than 100 individuals were uncovered in 52 graves, and the remains of 86 of them were preserved sufficiently well for skeletal [35] and ancient DNA analysis [13, 14].

The ancient Zayukovo-3 cemetery is in the middle part of the Baksan Gorge (Republic of Kabardino-Balkaria, Russia) in the foothill zone at an altitude of 900 – 1000 meters above sea level. It is located about 60 km southeast of the Klin-Yar III cemetery. Zayukovo-3 was an elite burial site for 1500 years (from the 8th century BC to the 7th century AD) [36].

There are several common features in the rites of Klin-Yar III and Zayukovo-3 cemeteries during the Koban culture period. Inhumations were laid out crouched, and the graves sometimes are stone-lined and/or have a stone cover.

Early stage of Alanic Culture description: “Bratskiye 1-ye Kurgany” barrow cemetery

The barrow cemetery of “Bratskiye 1-ye Kurgany” is located between Bratskoe village (Republic of Chechnya, Russia) and the administrative border with the Republic of North Ossetia–Alania (Russia). This archeological site was first studied in 1963, and again in 1971. Eleven underground catacombs of the Alanic culture of the 3rd–4th centuries AD were opened during this fieldwork [14, 33]. Rescue excavations in 2018 reported an additional 324 burials (barrows and flat graves) of the early stage of Alanic culture of the Middle Terek, dating to the first half of the 3rd century – first half of the 5th century AD [14, 33].

Archeological specimens and their origin

The materials used in the study were obtained during archeological excavations of Koban culture burials (Supplementary Table 1) as described above and in previous studies [13, 21, 36]. The present study uses five Koban culture individuals (four from the site of Zayukovo-3 near the village of Zayukovo and one from the Klin-Yar III cemetery near modern Kislovodsk), with the highest percentage of endogenous DNA and reference genome coverage that were sequenced previously [13], and one individual attributed to the early stage of Alanic period (“Bratskiye 1-ye Kurgany” barrow cemetery). Detailed information about the excavated graves and associated grave goods is presented in Supplementary Note 1.

Ancient DNA extraction, DNA library preparation, and DNA sequencing

Genetic material was extracted from the teeth in the clean rooms for ancient DNA extraction in National Research Center “Kurchatov Institute” (Moscow, Russia) using silica-bead-based DNA isolation protocol [37] that contains five main steps:

1. Teeth drilling for powder;

2. Teeth powder digestion in buffer containing EDTA and proteinase K;

3. DNA enrichment using silica beads in the binding buffer, which contains tris(hydroxymethyl)aminomethane (TRIS), sodium acetate, sodium chloride, and guanidine thiocyanate;

4. DNA extract washing in ethanol;

5. DNA elution from the beads.

Each tooth was mechanically cleaned of soil residues and treated with ultraviolet (UV) light for 15 min before drilling. A separate drill was used to remove the enamel layer. Then, it was changed, and the dentin and pulp tissues were drilled out using a different UV-sterilized drill. As a result, depending on tooth size, each sample contained 400-700 mg of dentin and pulp tissue powder. Multiplexed DNA libraries were constructed using an Ovation® Ultralow Library System V2 kit (NuGEN, USA). Multiple negative controls were used during the ancient DNA extraction and DNA library preparation. Amplified DNA libraries were quantified using a high-sensitivity chip on a 2100 Bioanalyser instrument (Agilent Technologies, USA) and sequenced on the Illumina Novaseq6000 system (Illumina, USA). The S2 flowcell of Illumina Novaseq6000 genome analyzer was used for the sequencing of six DNA libraries with paired-end reads of 150 bp in length.

Ancient DNA analysis

Raw sequencing data were converted to FASTQ format with bcl2fastq2 (v2.17). DNA reads of each library were processed through PALEOMIX 1.3.2 pipeline [38]. We used AdapterRemoval v2 [39] for the trimming of library adapters and low-quality bases (BaseQ <5 or Ns) and for the elimination of short (less than 25 bp) or non-informative (more than 30 bp of missing data) DNA reads. Moreover, AdapterRemoval v2 merged paired-end reads before the alignment stage (Supplementary Table 5). Filtered reads were mapped against the human reference genome (GRCh37/hg19) using BWA-MEM v0.6.2 [40] with option “FilterUnmappedReads: yes”. PCR duplicates were trimmed with Picard v1.128 (https://broadinstitute.github.io/picard), and the percentage of endogenous DNA for each library was estimated as the ratio between the number of mapped and the number of post-filtering reads (Supplementary Table 2). Postmortem DNA damage patterns were analyzed using MapDamage v2.0 [41]. Three nucleotides from each end of each DNA read were trimmed using trimBam from bamUtil repository [42] to reduce the impact of nucleotide misincorporations (changes from C → T to G → A base pairs) in the subsequent analyses. Pseudo-haploid calls were generated for randomly selected reads for each human specimen from the targeted 1,233,013 SNPs in genome using pileupCaller genotype caller (https://github.com/stschiff/sequenceTools).

Mitochondrial DNA contamination test was carried out with a Bayesian approach as described by Fu and colleagues [43] and using schmutzi contamination test [44]. Male X chromosome contamination was estimated using contamination.R code from ANGSD [45]. We used methods of moments (MoM) applied and maximum likelihood (ML) estimates from the “Method 2” parameter.

Following the methodology described by Skoglund and colleagues [46], we performed anthropological sex determination by assessing the ratio of Y chromosome reads to the total number of reads mapping to both sex chromosomes. We utilized the hapCon tool [20] to carry out a contamination test. To refine the data, we employed samtools [47] to filter the bam files, specifically targeting sequences with a mapping quality of ≥30 and bases with a base quality of ≥30. For the hapCon analysis, we employed both the 1240k Reference Panel and the 1000 Genome Reference Panel as reference databases.

We determined the Y chromosome haplogroups of male individuals by analyzing the captured SNPs on their Y chromosome. We used the dataset provided by the International Society of Genetic Genealogy (ISOGG) for this analysis. During the process, we ignored C-to-T transitions on the forward reads and G-to-A transitions on the reverse reads. To identify the Y chromosomal haplogroups, we referred to the phylogenetic tree available in ISOGG version 15.73 (accessible at http://www.isogg.org/tree). By examining this tree, we identified the most ancient allele downstream and the most recent allele upstream to determine the Y-chromosome haplogroups. Contamination and sex determination metrics are presented in Supplementary Tables 3–5.

The Allen Ancient DNA Resource (AADR) database (https://reich.hms.harvard.edu/allen-ancient-dna-resource-aadr-downloadable-genotypes-present-day-and-ancient-dna-data) was selected as a reference panel of genotypes of modern and ancient people, since ancient and modern individuals were genotyped in 1,233,013 sites. We combined our newly generated ancient samples with the modern and ancient populations from the publicly available AADR dataset.

Principal Component Analysis (PCA)

The smartpca tool from the EIGENSOFT package [48] was used for principal component analysis (PCA) for combined datasets:

1. Koban, Alanic, Sarmatian, Scythian, and Cimmerian culture bearers (Supplementary Dataset 1: for Fig. 3 sheet);

2. Bearers of ancient Caucasian cultures (Eneolithic, Bronze, and Iron Ages) and archeological cultures from the adjacent European steppe, e.g., Sarmatians, Scythians, and Cimmerians (Supplementary Dataset 1: for Fig. S7 sheet);

3. Ancient Caucasian culture bearers and individuals from the modern Caucasian ethnic groups (Supplementary Dataset 1: for Fig. S8 sheet);

4. Ancient Caucasian culture bearers (Eneolithic, Bronze, and Iron Ages), bearers of archeological cultures from the adjacent European steppe (Sarmatians, Scythians, and Cimmerians), and modern Western and Eastern Eurasian individuals (Supplementary Dataset 1: for Fig. S9 sheet).

In all the PCA plots we generated; projections to modern human populations (Supplementary Dataset 1: for Projection) with HO panel containing 412 thousand SNPs were used. Modern datasets are not shown on several PCA plots (e.g., Fig. 3; Fig. S7) to preserve their scale and readability. We used the following default parameters for smartpca analysis: lsqproject: YES; numoutlieriter: 0; shrinkmode: YES. The final list of populations and individuals is presented in Supplementary Dataset 1.

f-statistics and admixture analysis

In all the ADMIXTURE plots we generated; projections of modern human populations (Supplementary Dataset 1: for Projection) were used. The qp3Pop tool with default parameters (ADMIXTOOLS package [48]) was used for f3-statistics analysis. The qpDstat tool from the same package was used for f4-statistics analysis. The genome wide dataset was filtered in PLINK [49] with parameter indep-pairwise 200 25 0,4 which retained 436,763 SNPs for the Human Origins dataset. After that, we carried out admixture analysis with ADMIXTURE tool (v.1.23) [50] with default cv = 5 cross-validation parameter, and the number of ancestral populations ranged between K  =  3 and K  =  12 in 100 bootstraps, as described previously [3]. Сross-validation error rates showed that the least error rate was produced by K  =  12. The combined datasets of individuals from ancient Caucasian cultures (Eneolithic, Bronze, and Iron Ages), bearers of archeological cultures from the adjacent European steppe (Sarmatians, Scythians, and Cimmerians), and reference Neolithic specimens of Anatolia (Anatolia_N), Caucasus hunter-gatherers (CHG), and Yamnaya culture were included in the Admixture analysis. The populations for f-statistics analyses were selected from the same dataset based on archeological context and the results of PCA and Admixture analyses. Here, we obtained ancient populations/individuals (Source 2) who made the most significant genetic impact on the Alanic culture individuals together with Koban culture bearers (Source 1). The final lists of populations and individuals used for ADMIXTURE and f-statistic analyses are presented in Supplementary Dataset 1.

Kinship analysis

READ [51] tool with default parameters was used for kinship analysis between five Koban and one early Alanic culture bearers using the same methodology as described in Wang et al. [3]. Kinship analysis was carried out based on SNPs detected in the previous step. Genotyping data for true unrelated as well as related individuals from the Allen Ancient DNA Resource (AADR) database were also included in the kinship analysis. The final list of individuals used in kinship analysis and READ estimates for the degree of kinship among individuals are included in Supplementary Dataset 2.

Author contributions

Conceptualization: F.S.S., S.M.R, D.S.K. and A.V.N.; data curation: F.S.S.; formal analysis: F.S.S. and S.M.R.; funding acquisition: D.S.K. and A.V.N.; investigation: E.S.B., S.V.T., N.V.S., A.B.B., H.H., A.A.K., S.V.D., V. Yu. M., T.Yu.S., M.V.D. and I.K.R.; methodology: E.S.B., S.V.T., N.V.S., A.B.B., H.H., A.A.K., S.V.D., T.Yu.S., M.V.D. and I.K.R.; project administration: D.S.K. and A.V.N.; resources: F.S.S. and S.M.R.; software: F.S.S. and S.M.R.; supervision: D.S.K. and A.V.N.; visualization: F.S.S., A.A.K. and A.V.N.; writing—original draft: F.S.S., D.S.K. and A.V.N.; writing—review and editing: all authors.

Data availability

The raw reads of the Koban and Alanic culture bearers generated in this study are available for download through the National Center for Biotechnology Information, BioProject ID PRJNA797283. The accession numbers of the previously published genomic data that were reprocessed in this study are available in the Supplementary Materials. All other data are included in the paper or are available upon request.

Competing interests

The authors declare no competing interests.

Ethical approval

This research was carried out under the ethics guidelines for DNA research on human remains described previously: Alpaslan-Roodenberg S, Anthony D, Babiker H, Bánffy E, Booth T, Capone P et al. Ethics of DNA research on human remains: five globally applicable guidelines. Nature 2021; 599: 41–46.

Supplementary information

The online version contains supplementary material available at 10.1038/s41431-023-01524-4.