Haplogroup R1a (Y-DNA)

Haplogroup R1a
Possible time of origin Less than 18,500 YBP[1]
Possible place of origin Eurasia, most probably South Asia, or Central Eurasia.
Ancestor R1 (R-M173)
Descendants R1a1a1 to R1a1a8.
Defining mutations M420 (also includes M17, M198, SRY1532.2 and SRY10831.2)
Highest frequencies Scandinavia, Central Europe, Eastern Europe, Central Asia, South Asia and Siberia. (See List of R1a frequency by population)

Haplogroup R1a is the phylogenetic name of a major clade of Human Y-chromosome DNA haplogroups. In other words, it is a way of grouping a significant part of all modern men according to a shared male-line ancestor. It is common in many parts of Eurasia and is frequently discussed in human population genetics and genetic genealogy. One sub-clade (branch) of R1a, currently designated R1a1a, is much more common than the others in all major geographical regions. R1a1a, defined by the SNP mutation M17, is particularly common in a large region extending from South Asia and Southern Siberia to Central Europe and Scandinavia.[2]

Currently, the R1a family is defined most broadly by the SNP mutation M420. The recent discovery of M420 resulted in a reorganization of the known family tree of R1a, in particular establishing a new paragroup (designated R1a*) for the relatively rare lineages which are not in the R1a1 branch leading to R1a1a.

R1a and R1a1a are believed to have originated somewhere within Eurasia, most likely in the area from Eastern Europe to South Asia. Several recent studies have proposed that South Asia is the most likely region of origin. But on the other hand, as will be discussed below, some researchers continue to treat modern Indian R1a as being largely due to immigration from the Central Eurasian steppes.

Contents


Different meanings of "R1a"

Further information: Conversion table for Y chromosome haplogroups

The naming system commonly used for R1a remains inconsistent in different published sources, and requires some explanation.

In 2002, the Y chromosome consortium (YCC) proposed a new naming system for haplogroups, which has now become standard.[3] In this system, names with the format "R1" and "R1a" are "phylogenetic" names, aimed at marking positions in a family tree. Names of SNP mutations can also be used to name clades or haplogroups. For example, as M173 is currently the defining mutation of R1, R1 is also R-M173, a "mutational" clade name. When a new branching in a tree is discovered, some phylogenetic names will change, but by definition all mutational names will remain the same.

The widely occurring haplogroup defined by mutation M17 was known by various names, such as "Eu19",[4] in the older naming systems. The 2002 YCC proposal assigned the name R1a to the haplogroup defined by mutation SRY1532.2. This included Eu19 (i.e. R-M17) as a subclade, so Eu19 was named R1a1.[5] The discovery of M420 in 2009 has caused a reassignment of these phylogenetic names.[2][6] R1a is now defined by the M420 mutation: in this updated tree, the subclade defined by SRY1532.2 has moved from R1a to R1a1, and Eu19 (R-M17) from R1a1 to R1a1a.

Contrasting family trees for R1a
2002 Scheme proposed in YCC (2002) 2009 Scheme as per Underhill et al. (2009)
As M420 went undetected, M420 lineages were classified as either R1* or R1a (SRY1532.2, also known as SRY10831.2)
R1
 M173  
R1*

 All cases without M343 or SRY1532.2 (including a minority M420+ cases)

R1a
 SRY1532.2 
  (SRY10831.2)  

R1a* 

 
R1a1
 M17, M198 

 R1a1*

 M56 

 R1a1a

 M157 

 R1a1b

 M87, M204
M64.2

 
 R1a1c



R1b
M343

 sibling clade to R1a


After 2009, a new layer was inserted covering all old R1a, plus its closest known relatives
R1
 M173  
R1*

 All cases without M343 or M420 (smaller than old "R1a*")

R1a 
M420 

  R1a* All cases with M420 but without SRY1532.2

R1a1 
SRY1532.2 


  R1a1*(Old R1a*)


 R1a1a 
 M17, M198 

R1a1a*

M56
 

R1a1a1

M157
 

R1a1a2

 M64.2,..
 

R1a1a3

P98
 

R1a1a4

PK5
 

R1a1a5

M434
 

R1a1a6

 M458 
 

 R1a1a7*

 
M334 
 

 R1a1a7a


 Page68[7]
 

R1a1a8




R1b
M343

 Sibling clade to R1a (same as before)



Phylogeny

The R1a family tree now has three major levels of branching, with the largest number of defined subclades within the dominant and best known branch, R1a1a (which, as has been noted, will be found with various names; in particular, as "R1a1" in relatively recent but not the latest literature.)

Roots of R1a

Haplogroup R family tree
 
 Haplogroup R  
  Haplogroup R1  
M173
  M420 

  R1a

  M343 

 R1b

?

R1*



 Haplogroup R2


R1a, distinguished by several unique markers including the M420 mutation, is a subclade of Haplogroup R1, which is defined by SNP mutation M173. Besides R1a, R1 also has the subclades R1b, defined by the M343 mutation, and the paragroup R1*. There is no simple consensus concerning the places in Eurasia where R1, R1a or R1b evolved.

R1a (R-M420)

R1a, defined by the mutation M420, has two branches: R1a1, defined by the mutation SRY1532.2, which makes up the vast majority; and R1a*, the paragroup, defined as M420 positive but SRY1532.2 negative. (In the 2002 scheme, this SRY1532.2 negative minority was one part of the relatively rare group classified as the paragroup R1*.) Mutations understood to be equivalent to M420 include M449, M511, M513, L62, and L63.[2][6]

Only isolated samples of the new paragroup R1a* have been found by Underhill et al., mostly in the Middle East and Caucasus: 1/121 Omanis, 2/150 Iranians, 1/164 in the United Arab Emirates, and 3/612 in Turkey. Testing of 7224 more males in 73 other Eurasian populations showed no sign of this category.[2]

R1a1 (R-SRY1532.2)

R1a1 is currently defined by SRY1532.2, also referred to as SRY10831.2. SNP mutations understood to be always occurring with SRY1532.2 include M448, M459, and M516.[2] This family of lineages is dominated by the very large and well-defined R1a1a branch, which is positive for M17 and M198. The paragroup R1a1* (old R1a*) is positive for the SRY1532.2 marker but lacks either the M17 or M198 markers.

The R1a1* paragroup is apparently less rare than R1* but still relatively unusual, though it has been tested in more than one survey. Underhill et al. for example report 1/51 in Norway, 3/305 in Sweden, 1/57 Greek Macedonians, 1/150 Iranians, 2/734 Ethnic Armenians, and 1/141 Kabardians.[2] While Sahoo et al. reported R1a*(new R1a1*) for 1/15 Himachal Pradesh Rajput samples.[8]

R1a1a (R-M17 or R-M198)

R1a1a (old R1a1) makes up the vast majority of all R1a over its entire geographic range. It is defined by SNP mutations M17 or M198, which have always appeared together in the same men so far. SNP mutations understood to be always occurring with M17 and M198 include M417, M512, M514, M515.[2]

Currently, R1a1a has eight subclades of its own defined by mutations, but the vast majority of the incidence has not yet been categorized and is therefore in the paragroup R1a1a*.

R1a1a subclades

Currently, of the eight SNP-defined subclades of R1a1a only R1a1a7 (now R1a1a1g) has significant frequencies. R1a1a7 (now R1a1a1g) is defined by M458 and was found almost entirely in Europe, and with low frequency in Turkey and parts of the Caucasus. Its highest frequencies were found in Central and Southern Poland, particularly near the river valleys flowing northwards to the Baltic sea.[2]

R1a1a7 (now R1a1a1g) has its own SNP-defined R1a1a7a subclade, defined by the M334 marker. However this mutation was found only in one Estonian man and may define a very recently founded and small clade.[2]

Relative frequency of R1a1a6 (R-M434) to R1a1a (R-M17)
Region People N R1a1a-M17 R1a1a6-M434
Number Freq. (%) Number Freq. (%)
 Pakistan  Baloch 60 9 15% 5 8%
 Pakistan  Makrani 60 15 25% 4 7%
 Middle East  Oman 121 11 9% 3 2.5%
 Pakistan  Sindhi 134 65 49% 2 1%
Table only shows positive sets from N = 3667 derived from 60 Eurasian populations sample, Underhill et al. (2009)

R1a1a3, defined by the M64.2, M87, and M204 SNP mutations, is apparently rare: it was found in 1 of 117 males typed in southern Iran.

[9]

R1a1a6, defined by M434, was detected in 14 people (out of 3667 people tested) all in a restricted geographical range from Pakistan to Oman. This likely reflects a recent mutation event in Pakistan.[2]

R1a1a STR clusters

Genetic genealogists looking at high accuracy STR (microsatellite) haplotypes (as used in genealogy) have also identified clusters of similar within R1a1a. Such clusters equate to groups with probable common ancestry, but with no known SNP defining them yet.

Gwozdz (2009) has identified two clusters within R1a1a7 (now R1a1a1g) ("P" and "N"). Western Slavic cluster (L260 - R1a1a1g2 ) Cluster P was originally identified by Pawlowski (2002) and apparently accounts for about 8% of Polish men, making it the most common clearly identifiable haplotype cluster in Poland. Outside of Poland it is less common. Central European cluster (M458 - R1a1a1g) Cluster N is not concentrated in Poland, but is apparently common in many Slavic areas i.e Czech republic, Slovakia, Poland, Eastern Germany - Lusatia, Bavaria (former Bavaria Slavica, Western Ukraine. Archeologically M458 was found in the 14th century graves discovered on Usedom (Germany).

There is also M334 - R1a1a1g1 - one person in Estonia. Gwozdz also identified at least one large cluster of R1a1a* (not having M458), referred to as cluster K. This cluster is common in Poland but not only there. Northern European cluster (L265 - R1a1a1i2) Example is cluster G or L365 - R1a1a1i2 (also called Pomeranian) popular in Pomerania, Baltic coast of Germany (former Abodrite and Vieleti areas), former Eastern Prussia, Lithuania.

Klyosov (2009) notes a potential clade identified by a mutation on the relatively stable STR marker DYS388 (to an unusual repeat value of 10, instead of the more common 12), noting that this "is observed in northern and western Europe, mainly in England, Ireland, Norway, and to a much lesser degree in Sweden, Denmark, Netherlands and Germany. In areas further east and south that mutation is practically absent".

Both Gwozdz and Klyosov also note frequent close STR matching between part of the Indian R1a1a population, and part of the Russian and Slavic R1a1a population, indicating apparent links between these populations in a time-frame more recent than the age of R1a1a overall.

Distribution of R1a1a (R-M17 or R-M198)

Further information: List of R1a frequency by population
Further information: Y-DNA haplogroups by ethnic groups

R1a has been found in high frequency at both the eastern and western ends of its core range, for example in India and Tajikistan on the one hand, and Poland on the other. Throughout all of these regions, R1a is dominated by the R1a1a (R-M17 or R-M198) sub-clade.

South Asia

In South Asia R1a1a has often been observed with high frequency in a number of demographic groups. The main two subclades of R1a1a are R1a1a* and R1a1a7. R1a1a7 is positive for M458 an SNP that separate it from the rest of R1a1a. It is significant because M458 is a European marker and the epicenter is Poland. R1a1a influence into India was not from Europe since the M458 marker is rare in India.[8][10]

In India, high percentage of this haplogroup is observed in West Bengal Brahmins (72%) [10] to the east, Konkanastha Brahmins (48%) [10] to the west, Khatris (67%)[2] in north and Iyenger Brahmins (31%) [10] of south. It has also been found in several South Indian Dravidian-speaking Adivasis including the Chenchu (26%) and the Valmikis of Andhra Pradesh and the Kallar of Tamil Nadu suggesting that M17 is widespread in Tribal Southern Indians.[11]

Besides these, studies show high percentages in regionally diverse groups such as Manipuris (50%)[2] to the extreme North East and in Punjab (47%)[11] to the extreme North West.

In Pakistan it is found at 71% among the Mohanna of Sindh Province to the south and 46% among the Baltis of Gilgit-Baltistan to the north.[2] While 13% of Sinhalese of Sri Lanka were found to be R1a1a (R-M17) positive.[11]

Hindus of Terai region of Nepal show it at 69%.[12]

Europe

In Europe, R1a, again almost entirely in the R1a1a sub-clade, is found at highest levels among peoples of Eastern European descent (Sorbs, Poles, Russians and Ukrainians; 50 to 65%).[13][14][15] In the Baltic countries R1a frequencies decrease from Lithuania (45%) to Estonia (around 30%).[16] Levels in Hungarians have been noted between 20 and 60% [15][17]Rosser et al. (2000)Battaglia et al. (2008).

There is a significant presence in peoples of Scandinavian descent, with highest levels in Norway and Iceland, where between 20 and 30% of men are in R1a1a.[18][19] Vikings and Normans may have also carried the R1a1a lineage westward; accounting for at least part of the small presence in the British Isles.[20][21][22][23] In East Germany where Haplogroup R1a reaches a peak frequency in Rostock at a percentage of 31.3% of German males, it averages between 20%-30%.[24]

Haplogroup R1a1a was found at elevated levels amongst a sample of the Israeli population who self-designated themselves as Ashkenazi Jews, possibly reflecting gene flow into Ashkenazi populations from surrounding Eastern European populations, over a course of centuries. This haplogroup finding was apparently consistent with the latest SNP microarray analysis which argued that up to 55 percent of the modern Ashkenazi genome is specifically traceable to Europe.[25][26] Ashkenazim were found to have a significantly higher frequency of the R-M17 haplogroup Behar reported R-M17 to be the dominant haplogroup in Ashkenazi Levites (52%), although rare in Ashkenazi Cohanim (1.3%) and Israelites (4%).[14]

In Southern Europe R1a1a is not common amongst the general population, but it is widespread in certain areas. Significant levels have been found in pockets, such as in the Pas Valley in Northern Spain, areas of Venice, and Calabria in Italy.[27] The Balkans shows lower frequencies, and significant variation between areas, for example >30% in Slovenia, Croatia and Greek Macedonia, but <10% in Albania, Kosovo and parts of Greece.[15][28][29]

The remains of a father and his two sons, from an archaeological site discovered in 2005 near Eulau (in Saxony-Anhalt, Germany) and dated to about 2600 BCE, tested positive for the Y-SNP marker SRY10831.2.[30] The R1a1 clade was thus present in Europe at least 4600 years ago, in association with one site of the widespread Corded Ware culture.[31]

Central and Northern Asia

R1a1a frequencies vary widely between populations within central and northern parts of Eurasia, but it is found in areas including Western China and Eastern Siberia. This variation is possibly a consequence of population bottlenecks in isolated areas and the movements of Scythians in ancient times and later the Turco-Mongols. High frequencies of R1a1a (R-M17 or R-M198; 50 to 70%) are found among the Ishkashimis, Khojant Tajiks, Kyrgyzs, and in several peoples of Russia's Altai Republic.[17][32][33] Although levels are comparatively low amongst some Turkic-speaking groups (e.g. Turks, Azeris, Kazakhs, Yakuts), levels are very high in certain Turkic or Mongolic-speaking groups of Northwestern China, such as the Bonan, Dongxiang, Salar, and Uyghurs.[32][34][35] R1a1a is also found among certain indigenous Eastern Siberians, including:Kamchatkans and Chukotkans, and peaking in Itel'man at 22%.[36]

Middle East and Caucasus

R1a1a has been found in various forms, in most parts of Western Asia, in widely varying concentrations, from almost no presence in areas such as Jordan, to much higher levels in parts of Kuwait, Turkey and Iran.[37][38][39]

The Shimar (Shammar) Bedouin tribe in Kuwait show the highest frequency in the Middle East at 43%.[37]

Wells et al. (2001), noted that in the western part of the country, Iranians show low R1a1a levels, while males of eastern parts of Iran carried up to 35% R1a. Nasidze et al. (2004) found R1a in approximately 20% of Iranian males from the cities of Tehran and Isfahan. Regueiro et al. (2006), in a study of Iran, noted much higher frequencies in the south than the north.

Turkey also shows high but unevenly distributed R1a levels amongst some sub-populations. For example Nasidze et al. (2005) found relatively high levels amongst two Kurdish groups of Turkey, the Kurmanji (13%) and Zazaki (26%).[40]

Further to the north of these Middle Eastern regions on the other hand, R1a levels start to increase in the Caucasus, once again in an uneven way. Several populations studied have shown no sign of R1a, while highest levels so far discovered in the region appears to belong to speakers of the Karachay-Balkar language amongst whom about one quarter of men tested so far are in haplogroup R1a1a.[2]

Origins and hypothesized migrations of R1a1a

Most published discussions of R1a origins are actually about the origins of the R1a1a (R-M17 or R-M198) sub-clade, which is both numerically dominant, and the most studied part of R1a. Data so far collected indicates that there two widely separated areas of high frequency, one in South Asia, around Indo-Gangetic Plain, and the other in Eastern Europe, around Poland and Ukraine. The historical and prehistoric possible reasons for this are the subject of on-going discussion and attention amongst population geneticists and genetic genealogists, and are considered to be of potential interest to linguists and archaeologists also.

South Asian origin hypothesis

Several recent studies have found South Asia to have the highest level of diversity of Y-STR haplotype variation within R1a1a. On this basis, while several studies have concluded that the data is consistent with South Asia as the likely original point of dispersal (for example, Kivisild et al. (2003), Mirabal et al. (2009) and Underhill et al. (2009)) a few have actively argued for this scenario (for example Sengupta et al. (2005), Sahoo et al. (2006), Sharma et al. (2009), and Thangaraj et al. (2010). A survey study as of December 2009, including a collation of retested Y-DNA from previous studies, concluded that a South Asian R1a1a origin was the most likely proposal amongst the various uncertain possibilities.[2]

On the other hand, other recent studies such as Zhao et al. (2009) continue to treat R1a in modern India as being at least partly due to immigration from the northwest associated with Indoeuropean languages and culture. One argument for this, as stated for example by Thanseem et al. (2006), is that this is implied by the uneven distribution pattern of R1a between castes and regions. Higher castes and more northerly Indian populations are considered to be more directly descended from the populations who brought Indoeuropean languages to India, and they tend to have higher levels of R1a than lower castes, and more southerly populations, while tribal castes and non Indoeuropean speaking groups tend to have the lowest frequencies of R1a. In order to explain exceptions to this pattern, these authors propose that R1a in India is also partly due to earlier movements of people from central Asia.

Middle East

As mentioned above, R1a haplotypes are less common in most of the Middle East than they are in either South Asia or Eastern Europe or much of Central Asia. It has nevertheless been mentioned in speculation about the origins of the clade. This is both because there are above-described pockets of high frequency and diversity, for example in some parts of Iran and amongst some Kurdish populations. A Middle Eastern origin for R1a has long been considered a possibility, and is still considered to be consistent with known data.[2][9][11][15]

Central Eurasia

Cordaux et al. (2004) argued, citing data from 3 earlier publications, that R-M17 (R1a1a) Y chromosomes most probably have a central Asian origin.[41] Central Asia is still considered a possible place of origin by Mirabal et al. (2009) after their larger analysis of more recent data. However these authors also consider other parts of Asia, particularly South Asia, to likely places of origin. Zhao et al. (2009) also continues to treat central Asia as the most likely region of origin, mentioning that the low genetic diversity estimates for this region compared to India and Europe could be due to "a recent founder effect or drift that led to the high frequency of R1a in the Southeastern Central Asia".

Concerning R1a having a possible origin in the regions of central Asia or eastern Europe, there are two very different timeframes proposed by different researchers using different methods.

Early (pre-Holocene) eastern European migration hypotheses

Coalescent time estimates for R1a1a(xM458) STR from Underhill et al. (2009)
Location TD
W. India 15,800
Pakistan 15,000
Nepal 14,200
India 14,000
Oman 12,500
N. India 12,400
S. India 12,400
Caucasus 12,200
E. India 11,800
Poland 11,300
Slovakia 11,200
Crete 11,200
Germany 9,900
Denmark 9,700
UAE 9,700

A widely cited theory proposed in 2000 that there may have been two expansions: first, R1a1a originally spreading from a Ukrainian refugium during the Late Glacial Maximum; and then, the spread being magnified by the expansion of males from the Kurgan culture.[15] A recent survey argues that R1a1a could be old enough for this scenario, but find it more likely that it was initially in Asia even if it was in parts of Europe by approximately 11,000 years ago.[2]

Most age estimates for R1a1a having such an early presence in Europe come from papers using the "evolutionarily effective" methodology described by Zhivotovsky et al. (2004), the latest such example being Mirabal et al. (2009) and Underhill et al. (2009). Researchers using this dating method therefore conclude that any Neolithic or more recent dispersals of R1a1a do not represent the initial spread of the whole clade, and might be more visible in the distribution of a subclade or subclades. Underhill et al. (2009) remark on the "geographic concordance of the R1a1a7-M458 distribution with the Chalcolithic and Early Bronze Age Corded Ware (CW) cultures of Europe". However they also note evidence contrary to a connection: Corded Ware period human remains at Eulau from which Y-DNA was extracted of R1a haplogroup appear to be R1a1a*(xM458) (which they found most similar to the modern German R1a1a* haplotype.)

In papers where the Zhivitovsky method is not the only method used, Europe's R1a1a diversity is generally understood to have been shaped more significantly by more recent events, including not only the Bronze Age, but also the spread of Slavic languages. Dupuy et al. (2005) speculated that "R1a [in Norway] might represent the spread of the Corded Ware and Battle-Axe cultures from central and east Europe." Luca et al. (2006), looking at data from the Czech Republic suggested there was evidence for a rapid demographic expansion approximately 1500 years ago. Rebala et al. (2007) also detected Y-STR evidence of a recent Slavic expansion from the area of modern Ukraine. Gwozdz (2009) saw evidence for a "rapid population expansion somewhat less than 1,500 years ago in the area that is now Poland".

Later (Bronze age) Steppe culture hypothesis

Archaeologists recognize a complex of inter-related and relatively mobile cultures living on the Eurasian steppe, part of which protrudes into Europe as far west as Ukraine. These cultures from the late Neolithic and into the Iron Age, with specific traits such as Kurgan burials and horse domestication, have been associated with the dispersal of Indo-European languages across Eurasia. Nearly all samples from Bronze and Iron Age graves in the Krasnoyarsk area in south Siberia belonged to R1a1-M17 and appeared to represent an eastward migration from Europe.[42]

Geneticists believing that they see evidence of R1a1a gene-flow from the Eurasian Steppe to India have frequently proposed the involvement of these Steppe cultures in the process.[43] Such a Steppe origin for all or part R1a1a continues to be argued on the basis of DNA results from ancient remains from several South Siberian late Kurgan sites, including some from the Andronovo culture.[44] However, in recent discussions of this theory it is considered only to apply to a part of R1a1a, making this theory no longer incompatible with other origins theories for R1a more broadly defined.[2][45]

Popular science

Bryan Sykes in his book Blood of the Isles gives imaginative names to the founders or "clan patriarchs" of major British Y haplogroups, much as he did for mitochondrial haplogroups in his work The Seven Daughters of Eve. He named R1a1a in Europe the "clan" of a "patriarch" Sigurd, reflecting the theory that R1a1a in the British Isles has Norse origins. It should be noted that this does not mean that there ever was any clan or other large grouping of people, which was dominated by R1a1a or any other major haplogroup. Real clans and ethnic groups are made up of men in many Y Haplogroups.

See also

Evolutionary tree of Human Y-chromosome DNA (Y-DNA) haplogroups
most recent common Y-ancestor
A
A1b A1a-T
A1a A2-T
A2 A3 BT
B CT
DE CF
D E C F
G H IJK
IJ K
I J LT K(xLT)
L T M NO P S
O N Q R

Y-DNA by populations · Famous Y-DNA haplotypes

Notes

  1. ^ The Indian origin of paternal haplogroup R1a1* substantiates the autochthonous origin of Brahmins and the caste system
  2. ^ a b c d e f g h i j k l m n o p q r Underhill et al. (2009)
  3. ^ YCC (2002)
  4. ^ as used in Semino et al. (2000)
  5. ^ SRY1532.2 is also known as SRY10831.2
  6. ^ a b ISOGG phylogenetic tree
  7. ^ Also identifiable with the standardized SNP reference rs34351054.
  8. ^ a b Sahoo et al. (2006)
  9. ^ a b Regueiro et al. (2006)
  10. ^ a b c d Sengupta et al. (2005)
  11. ^ a b c d Kivisild et al. (2003)
  12. ^ Fornarino et al. (2009)
  13. ^ Balanovsky et al. (2008)
  14. ^ a b Behar et al. (2003)
  15. ^ a b c d e Semino et al. (2000)
  16. ^ Kasperaviciūte, Kucinskas & Stoneking (2005)
  17. ^ a b Tambets et al. (2004)
  18. ^ Bowden et al. (2008)
  19. ^ Dupuy et al. (2005)
  20. ^ Irish Heritage DNA Project, R1 and R1a
  21. ^ Passarino et al. (2002)
  22. ^ Capelli et al. (2003)
  23. ^ Garvey, D. "Y Haplogroup R1a1". Archived from the original on February 8, 2007. http://web.archive.org/web/20070208175917/http://freepages.genealogy.rootsweb.com/~dgarvey/DNA/hg/YCC_R1a1.html. Retrieved 2007-04-23. 
  24. ^ Significant genetic differentiation between Poland and Germany follows present-day political borders, as revealed by Y-chromosome analysishttp:[1]
  25. ^ Braya et al. (2010)
  26. ^ Nebel et al. (2001)
  27. ^ Scozzari et al. (2001)
  28. ^ Rosser et al. (2000)
  29. ^ Pericić et al. (2005)
  30. ^ The Ysearch number for the Eulau remains is 2C46S.
  31. ^ Haak et al. (2008)
  32. ^ a b Wells et al. (2001)
  33. ^ Kharkov et al. (2007)
  34. ^ Wang et al. (2003)
  35. ^ Zhou et al. (2007)
  36. ^ Lell et al. (2002)
  37. ^ a b Mohammad, T; Xue, Y; Evison, M; Tyler-Smith, C (2009). "Genetic structure of nomadic Bedouin from Kuwait". Heredity 103 (5): 425–33. doi:10.1038/hdy.2009.72. PMC 2869035. PMID 19639002. http://www.pubmedcentral.nih.gov/articlerender.fcgi?tool=pmcentrez&artid=2869035. 
  38. ^ Nasidze et al. (2004)
  39. ^ Nasidze et al. (2005)
  40. ^ http://www.zazaki.org/files/Kurds.pdf
  41. ^ Wells et al. (2001), Semino et al. (2000), and Quintana-Murci et al. (2001)
  42. ^ Keyser et al. (2009)
  43. ^ For several examples from 2002, see Semino et al. (2000), Passarino et al. (2001), Passarino et al. (2002) and Wells (2002)
  44. ^ See Keyser et al. (2009): 9 out of 10 male specimens were found to be in R1a1a, evidence felt by the authors to suggest that the Steppes Kurgan culture spread from Europe to Siberia.
  45. ^ Klyosov (2009)

References

Projects