Haplogroup R1a

[Kaiser]

Eki: Nepal should have a high concentration of R1a, since it is just next to the Indian state of Uttar Pradesh which is rich in R1a. Genetic matches of peoples so far apart geographically are intriguing, but possible. This possibility is more profound if we assume the place of origin to be somewhere between Nepal and Norway, with omnidirectional migrations resulting in the spread as we see it today. Poland & Ukraine fit in just fine. For Indian/Nepali R1a to spread out as far as Scandinavia (or vice versa) seems a bit far-fetched, unless there is a study to the contrary. Amazing stuff anyway!

[Eki]

I recently tested my autosomal DNA at DNATribes and curiously had about as high match with the Nepal Sherpas as with Finns, Icelandic and Norwegians, and also had some relatively high matches in northern India (see the attached file). First I thought the Asian matches were flukes, but now I found this curious mention on a Shetland DNA project page that says an R1a individual had an exact match in Nepal. Could it be that R1a people came from Nepal/Northern India to Scandinavia or the other way around?

http://www.davidkfaux.org/shetlandislandsY-DNA.html

"Conclusions: The ancestors of Lawrence Mathewson of Aywick and Utrabister, Mid Yell, son of Matthew Thomason and grandson of Thomas Mathewson (of Copister, South Yell, d. 1687), were likely Norse Vikings (suggested by R1a grouping) who settled in Shetland circa 800 AD. Due to patronymics, the surnames of his surviving descendants in the male line today are either Mathewson or Williamson. The DNA signature is very rare. In worldwide DNA databases, the largest number of close matches is with the Altai people of Central Asia. The only exact matches in about 100,000 samples are found in Nepal, and Western Norway, as well as Shetland including the large family with the surname Blance from Delting, whose relationship has not yet been ascertained via genealogical record sources."

[cacio]

Kaiser:

thanks for the information. The idea was to check if and to what extent diversity of R1a1 is due to a western/iranian origin (versus a local Indian origin or a northern route from the Hindukush).

I went and looked at the Regueiro paper myself - but it seems there aren't really enough observations to say much. Northern Iran has 33 total observations- southern iran 117, so really there's not much to say about N Iran. A striking thing is that the fraction of R1a is smaller in Iran than in India. Even in the south, it's 15%. In the north, it's hard to say - but there are only 2 observations out of 33. And the presence of R1axR1a1 (stressed in the paper)- it's just 1 observation in the North- so I don't see how one can use this as a sign of origin or higher diversity of R1a in Iran.

A more interesting thing though is the presence of R1b in numbers greater than in India (5-15%). This could point to a higher diversity of R1 in general in Iran. There are 3 undifferentiated R1 in S Iran (and one in the North) - these would be quite interesting if confirmed. I have not been reading about R1* but I can't recall its presence in any particular place (except in Africa/Cameroon).

So, anyway, it seems to me the Regueiro paper is suggestive for R1, but doesn't say much about R1a. It could have originated there, and from there to India. But the Indian numbers are probably more consistent with a migration from central asia as well (or primarily from). And one can even explain the Southern Iranian R1a as a migration from India into Iran. Further study may be helpful in isolating Central Asian vs Iranian haplotype.

Another suggestive piece of information in the paper is about my own L haplogroup. While much smaller in Iran, it seems distributed among all three subgroups L1-L2-L3 (though, again, we're talking about 5 observations). Whether this higher diversity is consistent with an Iranian origin rather than a NW Indian one remains to be seen.

cacio

[J Man]

The very high percentage of haplogroup J among the Parsis really does seem to point to a Persian origin.




Y-DNA: J2a*

[Kaiser]

Kaiser et al: is there any information on the variability of R1a in Iran, or in the relation between Indian and Iranian R1a? The variance contours move west to east, so one wonders what happens further west. cacio

Cacio: A recent paper, Iran: Tricontinental Nexus for Y-Chromosome Driven Migration, M. Regueiro et al, June 2006 (Human Heridity) addresses your R1a question: "The variance contours move west to east, so one wonders what happens further west?"

I quote: "A group R lineage of Asian descent, R1a1-M198, is found at higher frequencies in the south rather than the north of Iran (15.38 vs. 3.03%). Interestingly, the M198 lineage has been linked with the spread of the Kurgan culture originating in southern Russia/Ukraine in the region between the Black and Caspian Seas. Furthermore, Wells et al, described a reduction in frequency of M198 eastward across Siberia to the Altai-Sayan Mountains and Mongolia, and southward into India. This finding supports the inferred migration of the Indo-Iranians during the period 3,000 to 1,000 B.C. as proposed by Mallory, although models of one haplogroup coinciding with one migration are inherently simplistic, especiallywhen high frequency clades within contiguous continental areas are involved. In fact it has been shown that at least some R1a1 chromosomes dispersed earlier."

"From the disparate M198 frequencies observed for the north and south of Iran, it is possible to envision a movement southward towards India where the lineage may have had an influence on the populations south of the Iranian deserts and where the Dasht-e-Lut desert would have played a significant role in preventing the expansion of this marker to the north of Iran. The lower frequencies of M198 in the region of Anatolia (11.8% in Greece and 6.9% in Turkey, with a statistically significant longitudinal correlation) and the Caucasus (10% in Georgia, 6% in Armenia and 7% in Azerbaijan) suggests that population movement was southward towards India and then westward across the Iranian plateau. In addition, the detection of rare R1-M173* and R1a-SRY1532 lineages in Iran at higher frequencies than observed for either Turkey, Pakistan or India suggests the hypothesis that geographic origin of haplogroup R may be nearer Persia."

As for R1a in Pakistan, you were right in pointing out that the frequency spread map doesn't quite depict an idea of frequency as well as the actual tables do. So here are a few quick ones on R1a1 in Pakistan:

Sindhi: 52%; Pathan: 40%; Baluchi: 28%; Brauhi (Dravidian): 24%; Makrani (East African-origin ethnic group): 25%; Kalash (Caucasus-origin ethnic group): 20%.

Remember that some of these groups have very small populations (eg, Kalash number only 3,000 people), so the percentages are not necessarily reflective of the overall R1a1 numbers.

As for Parsis whose influx into India (Gujerat-Ahmedabad and Maharashtra-Mumbai) and Pakistan (Sindh-Karachi) was (intriguingly) largely male-mediated, have the following Y-Chromosome composition:

J1/J2a,b = 40%; R1b3 = 27%; L = 18%; R1a = 8%; E1/E2 (YAP) = 5%.

This post is a bit long-winded, but I thought it might be helpful.

[lgmayka]

Interesting, if R1a did spread from western India then I do not think that they would have originally been Indo-European speakers.

Proto-Indo-European, by the usual linguistic definition, is less than 10,000 years old:

http://en.wikipedia.org/wiki/Proto-Indo-European_language#When_was_PIE_spoken.3F

Since R1a now appears to be of greater age than that, it stands to reason that the original tribe of R1a was speaking either some antecedent of Proto-Indo-European, or conceivably something very different. We have no reliable means to reconstruct linguistic prehistory that far back.

[father_R2]

Indeed, the highest variance of R1a1 in Baluchistan home to couple of millions of Dravidian speakers(Brahui) is also problamatic.

[cacio]

Kaiser et al:

is there any information on the variability of R1a in Iran, or in the relation between Indian and Iranian R1a? The variance contours move west to east, so one wonders what happens further west. Also, I didn't check the paper, but is the highest frequency found in balochistan or is it more generally in sindh/karachi etc.? it's hard to say from the map. Karachi would have lots of groups (eg the Parsi) whose inclusion may be problematic for variance issues.

cacio

[J Man]

One interpretation of Sengupta's 2006 research is that R1a originated long ago in Pakistan, and promptly spread to western India. Much later, after R1a had spread to Central Asia and Ukraine, it also descended upon northern India.

If I read Sengupta correctly, R1a has higher percentages in northern India, but much larger STR variance (typically considered a sign of haplogroup age) in western India.

Interesting, if R1a did spread from western India then I do not think that they would have originally been Indo-European speakers.




Y-DNA: J2a*

[Kaiser]

Within South Asia, the highest frequency of R1a is in Southern Baluchistan (Pakistan) and Uttar Pradesh (India). On the other hand, the microsatellite variation is highest in Baluchistan, with a eastward clinal spread. Taken together, it can be surmised that at least in South Asia, Baluchistan is the original home of R1a.

I quote from Sengupta's paper: "Interestingly, within India, R1a1-M17, R2-124, and L1-M76 display considerable frequency and HG-associated microsatellite variance. The widespread geographic distribution of HG R1a1-M17 across Eurasia and the current absence of informative subdivisions defined by binary markers leave uncertain the geographic origin of HG R1a1-M17. However, the contour map of R1a1-M17 variance shows the highest variance in the northwestern region of India. (Baluchistan, as seen in Fig 4 of the paper: Spatial frequency and mean microsatellite variance distributions of Y-chromosome HGs in present-day India and Pakistan).

[lgmayka]

I thought that R1a was supposed to have originated in the area of the Ukraine?

One interpretation of Sengupta's 2006 research is that R1a originated long ago in Pakistan, and promptly spread to western India. Much later, after R1a had spread to Central Asia and Ukraine, it also descended upon northern India.

If I read Sengupta correctly, R1a has higher percentages in northern India, but much larger STR variance (typically considered a sign of haplogroup age) in western India.

[J Man]

Based on the scientific studies, R1a prevails in India's upper casts. It is rather doubtful that mostly India's upper casts have spread this haplogroup north west so significantly. It is more reasonable to assume vise versa, as fits the common opinion, that R1a has been brought into India. A fair amount of analyses has been done to prove this theory, and this is why it currently prevails.
This haplogroup has been spread by Indo-European speakers not just to India, but into eastern Europe, Scandinavia, and middle Asia. I originate from Ukraine (where my known ancestors used to live) carry this haplogroup and found quite a few full 12 markers matches with Scandinavian and Scottish (Shetland) folks. All of them carry R1a.

I thought that R1a was supposed to have originated in the area of the Ukraine? Isn't that what most of the geneticists have said? Maybe it is native to Eastern Europe and has been there for a very long time and spread out from there? You have to look for diversity for origins of haplogroups.



Y-DNA : J2a*

[sitw]

1Ra origin in India?

Table 1 on this article is quite impressive when it comes to overview of R1b (and also R1a) in different European populations.

Noaide

Based on the scientific studies, R1a prevails in India's upper casts. It is rather doubtful that mostly India's upper casts have spread this haplogroup north west so significantly. It is more reasonable to assume vise versa, as fits the common opinion, that R1a has been brought into India. A fair amount of analyses has been done to prove this theory, and this is why it currently prevails.
This haplogroup has been spread by Indo-European speakers not just to India, but into eastern Europe, Scandinavia, and middle Asia. I originate from Ukraine (where my known ancestors used to live) carry this haplogroup and found quite a few full 12 markers matches with Scandinavian and Scottish (Shetland) folks. All of them carry R1a.

[Stevo]

Table 1 on this article is quite impressive when it comes to overview of R1b (and also R1a) in different European populations.

Noaide

It tends to show what I noticed awhile ago, i.e., that high levels of R1a tend to correspond with satem-speaking Indo-European regions, while high levels of R1b correspond with centum-speaking Indo-European areas.

I was kind of surprised, however, at the fairly significant level of R1b in Croatia, which explains my slow-marker matches there. According to that table, R1b is pretty stout among the Czechs and Slovaks (around 35%). I have matches among those folks, too.

I also have a real taste for Pilsner Urquell beer, the original Czech Pilsner.

Maybe it's hereditary?

[lgmayka]

Are 12/12 matches possible across subclades?

Yes, that can happen. For example, J1 and J2 sometimes have the same 12 markers, and G and G2 sometimes can't be told apart at 12 markers either.