y - caucasus languages

8/2/2019 Y - Caucasus Languages

1/7

Abstract A previous analysis of mtDNA variation in theCaucasus found that Indo-European-speaking Armeniansand Turkic-speaking Azerbaijanians were more closelyrelated genetically to other Caucasus populations (whospeak Caucasian languages) than to other Indo-Europeanor Turkic groups, respectively. Armenian and Azerbaijan-ian therefore represent language replacements, possiblyvia elite dominance involving primarily male migrants, inwhich case genetic relationships of Armenians and Azer-baijanians based on the Y-chromosome should more closelyreflect their linguistic relationships. We therefore ana-lyzed 11 bi-allelic Y-chromosome markers in 389 malesfrom eight populations, representing all major linguisticgroups in the Caucasus. As with the mtDNA study, basedon the Y-chromosome Armenians and Azerbaijanians aremore closely-related genetically to their geographicneighbors in the Caucasus than to their linguistic neigh-

bors elsewhere. However, whereas the mtDNA results showthat Caucasian groups are more closely related geneticallyto European than to Near Eastern groups, by contrast theY-chromosome shows a closer genetic relationship withthe Near East than with Europe.

Introduction

The Caucasus, the region between the Caspian and BlackSeas, is characterized by extremely high linguistic diver-

sity with four major families (South Caucasian, NorthCaucasian, Indo-European, and Altaic) spoken by more then50 ethnic groups. In addition, a major geographic barrier,the Caucasus Mountain range, divides the region into theNorth and South Caucasus sub-regions. The Caucasustherefore offers the opportunity for studing the impact oflinguistic diversity and geographic barriers on the geneticstructure of human populations. In particular, the presenceof populations who are linguistically-related to popula-tions outside the Caucasus, such as the Armenians (whospeak an Indo-European language) and the Azerbaijanians(who speak a Turkic language) raises the question as towhether language affiliation or geographic proximity bestexplains the genetic relationships of these Caucasian pop-ulations.

We have previously addressed this question by study-ing sequence variation in the first hypervariable segment

(HV1) of the mtDNA control region and Alu insertionpolymorphisms in Caucasian populations (Nasidze andStoneking 2001; Nasidze et al. 2001). Both studies haveshown that Armenians and Azerbaijanians are more closelyrelated genetically to their geographic neighbors in theCaucasus and not to their linguistic neighbors elsewhere.In addition, both studies have demonstrated that Caucasianpopulations are genetically intermediate between Euro-pean and Near Eastern populations, but that they are moreclosely related to European than to Near Eastern popula-tions.

The genetic results thus suggest that both the Armen-ian and Azerbaijanian languages represent language re-

placements in the Caucasus. The origins of the Armenianlanguage are obscure, but the Azerbaijanian language wasprobably introduced in the 11 th century AD by centralAsian nomads (Johanson 1998). A common mechanismof language replacement is elite dominance (Renfrew1991), whereby the language of a small invading group isadopted by the larger resident population, either because itis imposed by force or because it is considered sociallydesirable to speak the language of the invaders. If the in-vading group is primarily male, then one might expectpatterns of Y-chromosome variation to retain some traceof the invaders. To test this hypothesis of language re-

Ivan Nasidze Tamara Sarkisian Azer Kerimov Mark Stoneking

Testing hypotheses of language replacement in the Caucasus:

evidence from the Y-chromosome

Hum Genet (2003) 112: 255261DOI 10.1007/s00439-002-0874-4

Received: 9 August 2002 / Accepted: 23 October 2002 / Published online: 14 December 2002

ORIGINAL INVESTIGATION

I. Nasidze () M. StonekingMax Planck Institute for Evolutionary Anthropology,Inselstrasse 22, 04103 Leipzig, GermanyTel.: +49-341-9952505, Fax: +49-341-9952555,e-mail: [email protected]

T. SarkisianCenter of Medical Genetics,National Academy of Sciences of Republic of Armenia,5/1 Zakyan Str., 375010 Yerevan, Armenia

A. KerimovScientific-research Institute of Haematology and Transfusiology,Azerbaijan Republic Ministry of Health,Gashgay Street 87, Baku, Azerbaijan

Springer-Verlag 2002


2/7

placement via elite dominance, we report here an analysisof Y-chromosomal bi-allelic markers in the same set ofCaucasian populations previously analyzed for mtDNAHV1 sequence variation (Nasidze and Stoneking 2001).

Materials and methods

Subjects

A total of 389 samples (289 whole blood and 100 cheek cellswabs) from unrelated male individuals were collected in the fol-lowing eight autochthonous populations (Fig. 1): Georgians (SouthCaucasian speakers), Armenians (Indo-European speakers), Azer-baijanians (Turkic speakers), and Abazinians, Kabardinians, In-gushians, Chechenians, and Darginians (North Caucasian speak-ers). DNA from the blood samples had been used previously foranalyses of mtDNA HV1 sequence variation and Alu insertionpolymorphisms (Nasidze and Stoneking 2001; Nasidze et al.2001); cheek cell swabs from additional males from Armenia andAzerbaijanwere collected in order to increase the sample size forY-chromosome studies. Informed consent and information aboutthe birthplace of the donor and the donors parents and grandpar-ents were obtained from all donors. Genomic DNA from bloodsamples was extracted by using an IsoQuick DNA extraction kit

(Orca Research, Bothell, Wash., USA) or a conventional phenol-chloroform method (Maniantis et al. 1982). DNA from cheek cellswabs was extracted by using a conventional salting-out procedure(Miller et al. 1988). Published Y-chromosome single nucleotidepolymorphism (Y-SNP) data from the South Caucasus were usedfrom 25 Svans (Georgia), 25 Georgians from Kazbegi (Georgia),12 Lezgi (Azerbaijan), and 17 South Ossetians (Wells et al. 2001).Additional published data on 21 Azerbaijanians and 47 Armenians(Wells et al. 2001), and 63 Georgians (Semino et al. 2000; Wellset al. 2001), did not differ in any respect from our samples (datanot shown) and were not included in the analysis to avoid weight-ing the results too heavily on these populations. Published Y-SNPdata (Semino et al. 2000; Wells et al. 2001) for European, NearEastern, and Central Asian populations were also included.

Molecular analysis

Ten SNP markers previously reported to be polymorphic in Europeand the Near East (Semino et al. 2000) were typed in all samples:RPS4Y (M130), M9, M89, M124, M45, M173, M17, M201, M170,and M172 (Underhill et al. 2000 and references therein); the YAPAlu insertion polymorphism (Hammer and Horai 1995) was alsotyped. For all SNP markers except M130 (RPS4Y), Taqman (Ap-plied Biosystems) assays were designed. Primers and dye-labeledprobes were designed by using Primer Express (Version1 for Mac-intosh Power PC; PE Biosystems). Primer and probe sequences aregiven in Table 1. Reaction mixes were prepared as described pre-viously (Morin et al. 1999). M130 was typed by using the poly-merase chain reaction/restriction fragment length polymorphismprocedure described elsewhere (Kayser et al. 2000), whereas theYAP Alu insertion was typed as described previously (Hammerand Horai 1995). All samples were typed for all markers; no dis-crepancy was found between the hierarchical order of markers de-scribed by Underhill et al. (2000) and our results. The Y-SNP hap-logroup nomenclature used here is according to the recent recom-mendations of the Y Chromosome Consortium (2002). The phylo-genetic relationship of theY-chromosome haplotypes, based on the11 bi-allelic markers analyzed here, is shown in Fig. 2.

Statistical analysis

Haplogroup diversity and F st values were calculated with Arlequin2.000 (Schneider et al. 2000), which was also used to carry outMantel tests for correlations between matrices. Multidimensionalscaling (MDS) analysis (Kruskal 1964) of the Fst values was car-ried out with STATISTICA (StatSoft). Programs in PHYLIP3.5c(Felsenstein 1993) were used to construct a neighbor-joining treebased on pairwise Fst values.

256

Fig.1 Map of the Caucasusindicating the Y-SNP haplo-group frequencies in European,Near Eastern, Caucasian, andCentral Asian populations (ABAbazinians,KA Kabardinians,

INIngushians, CHChecheni-ans,DA Darginians, SVSvans,OS South Ossetians,KGeor-gians from Kazbegi, GE Geor-

gians,AR Armenians,AZAzer-baijanians,LE Lezginians). SV,K, andLE are from Wells et al.(2001)


3/7

Results and discussion

Y-SNP haplogroups in the Caucasus

Eleven Y-SNP haplogroups (Fig. 1) were found in theCaucasus (Table 2, Fig. 1). The most frequent haplogroupswere F*, G* and J2*; together the frequency of these

three haplogroups was 0.530.86 in all groups except forthe Darginians. The Darginians had a high frequency ofhaplogroup I* (0.58), which elsewhere was found at a fre-quency of 0.1 or less. Svans were previously reported(Wells et al. 2001) to have a high frequency (0.92) of hap-logroup F*, and Kazbegi to have a high frequency of hap-logroup J2* (0.72). No other groups had any single haplo-group at a frequency greater than 0.5, with the exceptionof the small sample of 12 Lezgi, for which the frequencyof haplogroup F* was 0.58. The Darginians, Lezgi, Svans,and Kazbegi correspondingly had the lowest haplogroupdiversities (0.1530.652), whereas for the other groups,the haplogroup diversity was 0.7790.855.

Two of the three common Caucasus haplogroups (F*and J2*) are also common in Near Eastern populationsLebanese, Turks, Syrians (Semino et al. 2000), and Irani-ans (Wells et al. 2001), with average frequencies of 0.165and 0.28, respectively, but present in lower frequencies inEurope (average frequencies 0.021 and 0.074, respec-tively). The third common Caucasus haplogroup, G*, israre in Europe (frequency=0.061) and, in the Near East,has been reported only in the Turkish and Lebanesegroups (Semino et al. 2000). Haplogroup R1*, which iscommon in Western and Central Europe, is observed inthe South Caucasus at frequencies higher then 0.1, whereasin the North Caucasus, it is absent or nearly so. Haplo-

group R1a1*, which is observed in high frequencies in thesouth-west Mediterranean region, Eastern Europe, andCentral Asia, is present at a low frequency in the Cauca-sus, similar to the Near East. The other Caucasus Y-haplo-groups occur at low frequency.

The Darginian, Svan, and Kazbegi groups appear to beoutliers compared with the other Caucasus groups. The Fstvalue was highest between Svans and other Caucasus groups(average Fst= 0.332), followed by the Kazbegi (averageFst= 0.286) and Darginians (average Fst= 0.25), whereasthe average pairwise Fst value among the remaining Cau-casus groups was only 0.047. These high Fst values, cou-pled with the lower haplogroup diversity and reduced

number of haplogroups (Table 2) in the Darginians, Svans,and Kazbegi, are most likely the result of genetic drift op-erating in small isolated populations.

The correlation between the geographic and genetic(pairwise Fst) distances separating pairs of Caucasus pop-ulations was not statistically significant (Mantel test:Z=0.113,P =0.671). Removal of the outliers (Svans, Kaz-begi, and Darginians) resulted in a correlation that washigher but still non-significant (Z=0.301, P =0.134). Todetermine whether the Caucasus Mountains have an influ-ence on the genetic structure of Caucasian populations,we examined the correlation between geographic and ge-

257

Table1

Primerandprobesequenc

esforTaqManassayswithcorrespondin

gannealingtemperatures(italicnucleotideinpolymorphicsite)

SNP

Forwardprimer

Reverse

primer

Probes

a

Annealing

marker

temperature

M9

CCTGAAATACAGAACTGCAAAGAA

GACAT

TGAACGTTTGAACATGTCT

CCTA

AGATGGTTGAATCCTCTTTATTTTTCTTTAATTT

62C

CCTA

AGATGGTTGAATGCTCTTTATTTTTC

TTTAATTT

M124

AATGCAAATTCCTGGGC

AA

TTCCTATAAAGCAAAGTTGAGGTTG

AACA

GGGAAGTCGGTTTAATAATACTGAG

TTTGT

62C

AACA

GGGAAGTTGGTTTAATAATACTGAG

TTTGTG

M45

AACAGTAACTCTAGGAGAGAGGATATCAA

CTATCTCCTGGCCTGGACCT

TGAA

AAATTATAGATAGGCAAAAAGCTCC

TTCTG

62C

CAGT

GAAAAATTATAGATAAGCAAAAAGC

TCCTTCT

M173

TACTGTAACTTCCTAGA

AAATTGGAAATAA

CAGTTTTCCCAGATCCTGAAAA

CAAG

GGCATTTAGAACACTTTGTCATCTGT

TA

62C

AGGG

CATTTAGAACCCTTTGTCATCTGTT

M17

AAATCAGATTCTGTCTA

CTCACCAGAG

TCACA

AAAATAGTTTGGCCACTT

TTGCTGGTTGTTACGGGGTTTTTTTAA

60C

TTGCTGGTTGTTACGGG-TTTTTTTAAGTG

M89

CCTGGATTCAGCTCTCT

TCCT

CCAGC

AAAGGTAGCTGCAAC

TTATGTACAAAAATCTCATCTCTCACTTTG

CCTGA

58C

TTATGTACAAAAATCTTATCTCTCACTTTG

CCTGA

M201

GCAATAGTTACTACTTG

AGTTACTATATTAGTGCAA

CATCCTATCAGCTTCATCCAACA

TCCA

GTATCAACTGAGGGTTTTCGTAATAGGTACTTA

62C

ATCC

AGTATCAACTGAGGTTTTTCGTAATA

GGTACTTA

M172

GGATTTTTCATTTTTATCCCCC

CCAGG

TACAGAGAAAGTTTGGACT

AACC

CATTTTGATGCTTTACTTAAAAGGTC

TTC

60C

ACCC

ATTTTGATGCTTGACTTAAAAGGTC

M170

TTTCATATTCTGTGCATTATACAAATTACTATT

CCACA

CAAAAACAGGTCCTCA

AAAA

ATCATTGTTCATTTTTTTCAGTGTGG

GT

58C

AAAA

TCATTGTTCCTTTTTTTCAGTGTGGG

aAncestralprobeisshownonthefirstlineandthemutantprobeontheseco

nd


4/7

netic distance for the South and North Caucasus sepa-rately, but neither was statistically significant (North Cau-casus: Z=0.331,P =0.133; South Caucasus: Z=0.346,P =0.65). Moreover, the average Fst value between South andNorth Caucasus populations (excluding outliers) was0.075, similar to that among North Caucasus populations(0.096) and that among South Caucasus populations(0.040). Therefore, the Caucasus Mountains appear not tohave had a detectable influence on the genetic structure ofCaucasus populations; instead, genetic drift operating in

small isolated populations seems to have dominated thegenetic structure of Caucasus populations.

Comparison of Caucasian, European,and Near Eastern Y-haplogroups

The haplogroup diversity in the Caucasus (average value:0.797) is almost as high as that in Central Asia (averagevalue: 0.824) and the Near East (average value: 0.769)and is significantly higher (t -test,P =0.024) than the hap-logroup diversity in Europe (average value: 0.633). AnMDS plot and neighbor-joining tree based on Fst values

(Fig. 3A, B) split European populations into Western andEastern groups, as has been observed previously (Seminoet al. 2000), with Central Asian populations falling in be-tween the Western and Eastern European groups. TheCaucasus populations are intermingled with Near Easternpopulations.

These patterns have further been confirmed by the pair-wise Fst comparisons; the mean pairwise Fst value for theCaucasus vs Europe is 0.254, whereas the mean F st valuefor the Caucasus vs the Near East is 0.079, which is sig-nificantly lower (t -test based on average Fst values jack-nifed over populations, P


5/7

average pairwise Fst value between Basques and Caucasiangroups is much higher (Fst = 0.563) than that betweenBasques and Indo-European groups (Fst= 0.311). The Cau-casus groups and Basques are not clustered together in ei-ther the neighbor-joining tree or the MDS plot (Fig. 3A, B).These results are in agreement with previous studies basedon classical markers and mtDNA HV1 sequence data(Bertorelle et al. 1995; Nasidze and Stoneking 2001).

Language replacements and genetic relationshipsin the Caucasus

The presence, in the Caucasus, of groups whose geo-graphic neighbors are not their linguistic neighbors allowsus to address the question as to which better explains thegenetic relationships of these groups: geography or lan-guage? In particular, Azerbaijanians speak a Turkic lan-guage but are surrounded by non-Turkic speakers, and Ar-menians speak an Indo-European language but are sur-rounded by non-Indo-European speakers. Previous mtDNA

analyses have shown that both Azerbaijanians and Arme-nians are more closely related genetically to other Cauca-sus groups than to their respective linguistic neighbors(Nasidze and Stoneking 2001), which indicates that theAzerbaijanian and Armenian languages were introducedvia language replacements. We have hypothesized thatthese language replacements have occurred via elite dom-inance (Renfrew 1991) in which a resident populationadopts the language of a small group of migrants either byforce or because speaking the language confers socialprestige. If such were the case, and the migrants were pri-marily male, then Y-DNA analyses may more closely re-flect the linguistic relationships of Azerbaijanians and Ar-

menians than have mtDNA analyses.Indeed, the Y-SNP haplogroups give similar results. TheTurkic-speaking Azerbaijanians are genetically closer totheir geographic neighbors in the Caucasus (averageFst=0.047) than to other Turkic-speaking groups (averageFst=0.105; t-test,P


6/7

Turkic groups, respectively. This may still reflect an elitedominance scenario, as presumably the original Indo-Eu-ropean/Turkic migrant groups were very small and/or didnot mix extensively with the resident groups. In anyevent, the migrant groups had a negligible genetic impact

on the resident groups. Currently, there are approximately8 million Azerbaijanians and 3.5 million Armenians, at-testing to the remarkable rapidity of these language shifts;such large-scale shifts would have to be accomplished bycultural rather than merely biological means.

260

Fig.3A, B The phylogeneticrelationship of Y-chromosomehaplogroups, based on 10Y-SNPs and the YAP marker.A MDS plot based on pairwiseFst values, showing relation-ships among the Caucasus, Eu-ropean, Near Eastern, and Cen-tral Asian populations (openstars with boldface italic popu-

lation names Caucasus groups,closed circles European popu-lations, open squares CentralAsian groups, closed diamondsNear Eastern populations). Thestress value for the MDS plotis 0.134. B Neighbor-joiningtree based on pairwise Fst val-ues for the same populations(boldface Caucasus groups)


7/7

However, the Y-haplogroups contrast sharply with theprevious mtDNA and Alu insertion polymorphism studies(Nasidze and Stoneking 2001; Nasidze et al. 2001) interms of the overall relationships of Caucasus popula-tions. For the Y-chromosome, the Caucasus populationsare more closely related to Near Eastern populations thanto European populations. Evidence for this is provided bythe higher diversity values and by the pairwise Fst values

and MDS plot. In particular, the Caucasus appears to be abreak zone in the Y-chromosome genetic landscape ofEurasia; haplogroups such as R1a1* are at a high fre-quency in East Europe but at low frequency in the Cauca-sus and the Near East, whereas haplogroups such as F*and J2* are common in the Caucasus and the Near East,but rare in Europe (Table 2, Fig. 1). By contrast, bothmtDNA and Alu insertion polymorphisms group Cauca-sus populations with Europe rather than the Near East.

How can these contrasting views on the relationship ofCaucasus populations be reconciled? A possible explana-tion is that the Y-chromosome results reflect repeated in-vading migrations from the Near East, migrations that

probably involved mostly males. Various Near Easterngroups invaded the Caucasus numerous times during thelast two millennia including the occupation of Georgia byArab caliphs after 654 AD (Muskhelishvili 1977), theSeljuc Turks invasion of the South Caucasus in the 11 th

century (Muskhelishvili 1977), and repeated invasions byTurks and Persians (Muskhelishvili 1977 and referencestherein). Indeed, strong Y-chromosomal gene flow fromthe Near East to the Caucasus has been clearly demon-strated for the interpolated maps of admixture proportionsin a recent study of Central Asian populations (Zerjal et al.2002). Although these male-mediated invasions from theNear East might possibly explain why the Caucasian Y-hap-

logroups are of Near Eastern origin, they do not explainwhy mtDNA and Alu insertion polymorphisms group theCaucasus with Europe. A possible explanation for this lat-ter result would be a common ancestry of Caucasian andEuropean populations. This hypothesized common ances-try could date back to pre-Neolithic times, as suggested byRenfrew (1992) who considers Caucasian languages to re-flect human dispersal before 15,000 years ago. On theother hand, it could reflect a route for Neolithic farmersfrom the Near East to Europe via the Caucasus; there areseveral securely dated Neolithic sites in the Caucasus thatare 6,0007,000 years old (Masson and Merpert 1982;Muskhelishvili 1977). Analyses of additional autosomalmarkers are needed to verify this hypothesized commonancestry of Caucasian and European populations.

Acknowledgements We thank Manfred Kayser for useful discus-sion, Peter Underhill for control DNA samples, and HiltrudSchaedlich for technical assistance. This research was supportedby funding from the Max Planck Society.

References

Bertorelle G, Bertranpetit J, Calafell F, Nasidze I, Barbujani G(1995) Do Basque-speaking and Caucasian-speaking populationsshare non-Indo-European ancestors. Eur J Hum Genet 3:256263

Felsenstein J (1993) PHYLIP (phylogeny inference package) v.3.5c. Department of Genetics, University of Washington, Seattle

Gamkrelidze T, Ivanov V (1995) Indo-European and the Indo-Eu-ropeans. De Gruyter, Berlin

Hammer M, Horai S (1995) Y chromosomal DNA variation andthe peopling of Japan. Am J Hum Genet 56:951962

Johanson L (1998) The history of Turkic. In: Johanson L, Csato E(eds) The Turkic languages. Routledge, London, pp8183

Kayser M, Brauer S, Weiss G, Underhill PA, Roewer L, Schiefen-hover W, Stoneking M (2000) Melanesian origin of Polynesian

Y chromosome. Curr Biol 10:12371246Kruskal JB (1964) Multidimensional scaling by optimizing good-ness of fit to a nonmetric hypothesis. Pyschometrika 29:127

Maniatis T, Fritsh EF, Sambrook J (1982) Molecular clonning: alaboratory manual. Cold Spring Harbor Laboratory, Cold SpringHarbor, NY

Masson VM, Merpert NJ (1982) Archaeology of USSR: Neolith ofUSSR. Nauka, Moscow, pp93164

Miller SA, Dykes DD, Polesky HF (1988) A simple salting outprocedure for extracting DNA from human nucleated cells. Nu-cleic Acids Res 16:1215

Morin PA, Saiz R, Monjazeb A (1999) High-throughput single nu-cleotide polymorphism genotyping by fluorescent 5 exonucle-ase assay. BioTechniques 3:538552

Muskhelishvili D (1977) The main problems of Georgian histori-calgeography. Metsniereba, Tbilisi

Nasidze I, Stoneking M (2001) Mitochondrial DNA variation andlanguage replacements in the Caucasus. Proc R Soc Lond [Biol]268:11971206

Nasidze I, Risch GM, Robichaux M, Sherry ST, Batzer MA,Stoneking M (2001) Alu insertion polymorphisms and the ge-netic structure of human populations from the Caucasus. Eur JHum Genet 9:267272

Renfrew C (1991) Before Babel: speculations on the origins of lin-guistic diversity. Cambridge Archaeol J 1:1323

Renfrew C (1992) Archaeology, genetics, and linguistic diversity.Man 27:445478

Ruhlen M (1991) A guide to the worlds languages. Stanford Uni-versity Press, Stanford

Schneider S, Roessli D, Excoffier L (2000) Arlequin v. 2000: asoftware for population genetic data analysis. Genetics andBiometry Laboratory, University of Geneva, Geneva, Switzer-

landSemino O, Passarino G, Oefner PJ, Lin AA, Arbuzova S, BeckmanLE, De Benedictis G, et al (2000) The genetic legacy of Paleo-lithicHomo sapiens sapiens in extant Europeans: a Ychromo-some perspective. Science 290:11551159

Underhill PA, Shen P, Lin AA, Jin L, Passarino G, Yang WH,Kauffman E, Bonne-Tamir B, Bertranpetit J, Francalacci P, etal (2000) Y chromosome sequence variation and the history ofhuman populations. Nat Genet 26:358361

Weale ME, Yepiskoposyan L, Jager RF, Hovhannisyan N, Khudo-yan A, Burbage-Hall O, Bradman N, Thomas MG (2001) Ar-menian Y chromosome haplotypes reveal strong regional struc-ture within a single ethno-national group. Hum Genet 109:659674

Wells RS, Yuldasheva N, Ruzibakiev R, Underhill PA, Evseeva I,Blue-Smith J, Jin L, Su B, Pitchappan R, Shanmugalakshimi S,Balakrishnan K, et al (2001) The Eurasian heartland: a conti-nental perspective on Y-chromosome diversity. Proc Natl AcadSci USA 98:1024410249

Y Chromosome Consortium (2002) A nomenclature system for thetree of human Y-chromosomal binary haplogroups. GenomeRes 2:339348

Zerjal T, Wells RS, Yuldasheva N, Ruzibakiev R, Tyler-Smith C(2002) A genetic landscape reshaped by recent events: Y-chro-mosomal insights into Central Asia. Am J Hum Genet 71:466482

261

y - caucasus languages

Documents