Monday, December 8, 2014

R1b = Araratean?

To dovetail off the last paper about R1b's vanishing act, here's another one, different methodology, different results.  Yepiskoposyan was part of previous study, same topic, same results at Nature [here]

Hat tip Bernard [His Blog]

Hovhannisyan et al, 2014

So be on the lookout for a car full of academics driving off a cliff.  Here's another R1b Neolithic agricultural diffusion theory, and here's why it didn't happen [here].

What you are looking at above is the frequency and variance of R1b, J2 and G.  The variance of R1b of the Ceyhan and Seyhan rivers (or Capadocia) is probably the old Armenian kingdom, but this area was also substantially settled by the Armenian Highland during the Pottery Neolithic.

What's intriguing and unexpected is the lack of diversity they show in the Caucasus for any lineage.  That is opposite of anything I would have imagined for this mountainous region.

They have Haplogroup G hovering over the old Natufian Mesolithic, which is essentially where agriculture developed.  I guess that kind of makes sense!

Another interesting take-away is the tight concentration of R1b's diversity over a place where it is mostly a non-factor anymore.  The exception of course is Armenia itself. 

Background: The peopling of Europe and the nature of the Neolithic agricultural migration as a primary issue in the modern human colonization of the globe is still widely debated. At present, much uncertainty is associated with the reconstruction of the routes of migration for the first farmers from the Near East. In this context, hospitable climatic conditions and the key geographic position of the Armenian Highland suggest that it may have served as a conduit for several waves of expansion of the first agriculturalists from the Near East to Europe and the North Caucasus.

Results: Here, we assess Y-chromosomal distribution in six geographically distinct populations of Armenians that roughly represent the extent of historical Armenia. Using the general haplogroup structure and the specific lineages representing putative genetic markers of the Neolithic Revolution, haplogroups R1b1a2, J2, and G, we identify distinct patterns of genetic affinity between the populations of the Armenian Highland and the neighboring ones north and west from this area.

Conclusions: Based on the results obtained, we suggest a new insight on the different routes and waves of Neolithic expansion of the first farmers through the Armenian Highland. We detected at least two principle migratory directions: (1) westward alongside the coastline of the Mediterranean Sea and (2) northward to the North Caucasus.

Different waves and directions of Neolithic migrations in the Armenian Highland, Investigative Genetics, Anahit Hovhannisyan1, Zaruhi Khachatryan, Marc Haber, Peter Hrechdakian, Tatiana Karafet, Pierre Zalloua, Levon Yepiskoposyan, 2014 [Link]


  1. Interesting post (and blog), thanks. Out of interest, where is R1b peak diversity within Europe? Is it in Sardinia/N.Italy by any chance? Best regards.

    1. Thanks. This opinion on diversity seems to be fluxing quite a bit now. You could probably throw darts at the map.
      I think it's fair to say it's not clinal, at least not yet. Some of the older clades may overlay legs of the Beaker network.

  2. The highest L23 diversity is in Pashtuns, then polish, going by 10 strs. That part of the Near East is misinformative. There's European subclades, plus an Armenian diaspora community there.

  3. @ Chad - I have just read up on L23. High diversity is in 'Bashkirs', not Pashtuns, along with Kosovans, Poles, and Armenians.

    As I explained in a comment I made on the 'Haplogroup H & Modern Europe', Bashkirs have experienced gene flow from West Asia:-

    HLA Haplotype:
    [one example of many]

    United Arab Emirates [N=298] 4.80%
    Jordan [N=15,141] 3.86%
    [Data for Egypt unavailable]
    Morocco Settat Chaouya [N=98] 3.60%
    Tunisia [N=100] 3.00%
    Pakistan Mixed Sindhi [N=101] 3.00%
    **RUSSIA SOUTH URAL TATAR [N=135] 1.90%**
    **RUSSIA CHUVASH [N=82] 1.80%**
    Tunisia Ghannouch [N=82] 1.80%
    Morocco Pop. 2 [N=110] 1.59%
    Portugal Aveiro [N=5,933] 1.30%
    Turkey Pop.2 [N=228] 1.30%
    Tunisia Pop.3 [N=104] 1.20%
    Portugal Braganca [N=301] 1.20%
    Gaza Palestinian [N=165] 1.20%
    Portugal Leiria [N=1,847] 1.10%
    Portugal Porto [N=7,937] 1.00%
    Portugal Santarem [N=3,865] 1.00%
    Portugal Guarda [N=797] 1.00%
    *RUSSIA SOUTH URAL BASHKIR [N=146] 1.00%**
    etc. etc.

    The West Asian geneflow also applies to Kosovans and Armenians. Poles could have received West Asian geneflow from via the Balkans.

  4. Bashkirs don't have very high diversity. Those haplotypes probably aren't informative on R1b. Might be be Neolithic or arise independently.

    1. Bashkir ethnogenesis is historically attested and took place in the early Middle Ages. The elevated R1b in this area is probably a founder effect in patrilineally organized and inbred clans (of which there are seven). This is an isolated linguistically Turkic language family people that absorbed some people from a population akin to the linguistically Uralic Komi people. Today, they are Sunni Muslims. Prior to reaching their current home in the Northern Urals, they were found on the Danubian Plain and South Urals.

  5. 1. Pashtuns 2. Poland 3. Turkey 4. Romania 5. Bulgaria

  6. That part of west Asia has seen Anatolian speakers, Greeks, Romans, Armenian diaspora, and maybe more. There are L11's, L51's, and p312's.

  7. Chad, were you saying that Pashtuns have the highest diversity of R1b as a whole, or just the L23 clade? If so could you provide your reference, thanks.

    What I read was on Wikipedia:

    R1b1a2 (R-M269)
    "[..]Kosovo is notable in also having a high percentage of descendant L23* or L23(xM412) at 11.4% unlike most other areas with significant percentages of M269* and L23* except for Poland with 2.4% and 9.5% and the Bashkirs of southeast Bashkortostan with 2.4% and 32.2% respectively.[7] Notably this Bashkir population also has a high percentage of M269 sister branch M73 at 23.4%.[7] Five individuals out of 110 tested in the Ararat Valley, Armenia belonged to R1b1a2* and 36 to L23*, with none belonging to known subclades of L23.[37]"

  8. "Those haplotypes probably aren't informative on R1b.."

    Well I provided data from another type of genetic marker (HLA haplotypes) in order to illustrate gene flow from West Asia towards the Urals.
    The haplotype is at highest frequency (and diversity of recombinants, although I didn't provide the data) in North Africa and West Asia. In the Uralic region it is not equilibrated, indicating a relatively recent [definitely <20kya] asymmetric migration from the source region [North Africa or West Asia] towards the Circum-Uralic region.

    Since the discussion centred on where R1b originated from. I believe the answer is more likely North Africa or West Asia, while you believe it is Central Asia or Pashtuns. I can see from HLA haplotypes that the Uralic region, Caucasus, Pakistan, and Central Asia have *all* experienced some recent gene flow from the direction of Northern Africa/West Asia within the last 20kya.

    "That part of west Asia has seen Anatolian speakers, Greeks, Romans, Armenian diaspora, and maybe more. There are L11's, L51's, and p312's."

    But I can clearly observe, using a different type of genetic marker, far more evidence of gene flow *from* West Asia *into* Anatolia, Greece, Balkans, Italy, Caucasus etc. than I can the other way around.
    (BTW - HLA haplotypes are inherited in pairs, you get one from your mother and one from your father).
    Where an HLA haplotype is most equilibrated indicates a likely source, where it is in disequilibrium indicates a likely direction of asymmetric migration.

  9. Check Myers 2010. If R1b was in West Asia it would be found in Nelithic Europe. It's not and we have no obvious migrations from West Asia after around 5000 BCE. Then look at Mako, which is connected to German Beakers and Vucedol. We have over a 20% increase in WHG and 12% ANE, from Baden to Mako. R1b is not from West Asia.

  10. Pashtuns have the highest L23 diversity going by 10 strs. Central Asians have neolithic ancestry, like Europeans and North Africans.

  11. "If R1b was in West Asia it would be found in Nelithic Europe. It's not "

    How can we *know* that it's not found in Neolithic Europe?

  12. In any case:

    Emerging genetic patterns of the european neolithic: Perspectives from a late neolithic bell beaker burial site in Germany

    "Ancient DNA analyses of human remains from the Late Neolithic Bell Beaker site of Kromsdorf, Germany showed distinct mitochondrial haplotypes for six individuals, which were classified under the haplogroups I1, K1, T1, U2, U5, and W5, and two males were identified as belonging to the Y haplogroup R1b."

  13. On another subject, I find R1b-V88 to be quite interesting. V88 is found in North Cameroon, Sudan, North Africa, Sardinia, Corsica, Southern France, Egypt, Palestine, Jordan, Iran.

    Kind of similar to the distribution of the A*33:01-B*14:02-C*08:02 HLA haplotype.

    And also the 13913C lactase persistence allele (a rarer one found only in Italy, Cameroon, Sudan, Ethiopia, and Saudi Arabian Bedouins).

  14. Check that. I'll have to search for that other study again. Myers has Romania as the most diverse at 8 STRs, at 10 STRs it is Pashtuns as the most diverse, and Romanians second.

  15. This comment has been removed by the author.

  16. According to Mjost, the work he's done now shows L23 diversity is higher on the Western Black Sea coast, with a max age of 4300bce. Perfectly matching steppe cultures down the coast.

  17. "According to Mjost, the work he's done now shows L23 diversity is higher on the Western Black Sea coast, with a max age of 4300bce. Perfectly matching steppe cultures down the coast."

    If R1a and R1b were split by some geographic / climate border and it wasn't western Europe vs eastern Europe then Romania/Ukraine and the area around the western Black sea coast seems like the next most likely place.

    1. My guess is that R1a was North of the Caucasus and R1b was to the South.

  18. "What's intriguing and unexpected is the lack of diversity they show in the Caucasus for any lineage. That is opposite of anything I would have imagined for this mountainous region."

    The Caucasus are a genetic sink, not a genetic source. These mountains are a refuge for relict lineages and so should not show all that much diversity.