Monday, December 9, 2013

More East African mtDNA Charts

Below are more East African mtDNA bar graphs from the Hirbo Thesis, the complementary YDNA charts can be seen in this post, along with the Boattini paper featured here, this gives us a more complete picture of East African mtDNA with a reasonable amount of detail.

Google Visualization API has been having problems for the past couple of months, so the tool tips as well as other functionalities of Google charts may not work, this post will be updated if they fix some of these issues.

With respect to some of the data points, the populations labeled with a * had their total number of samples adjusted in order for the percentages shown in Table 3.4.1 to make sense, that is, Orma has been adjusted from 20 to 21, Marakwet from 22 to 23, Pokot from 39 to 38, San from 11 to 12 and Bamoun from 18 to 20.



3 comments:

  1. A most interesting dataset, thanks a lot, Etyopis.

    I find of particular interest the wild array of "rare" African lineages, which are sometimes ill understood, such as L4, L5, L6 or even L0 itself.

    ReplyDelete
  2. Not sure if you have already seen this, but there is a relatively new (open access) paper that details mtDNA L0, where they postulate a South African origin for the lineage, which they make a good case for in terms of pyhlogenetic inference, but they obviously lack the archaeological proof for their proposal - http://www.plosone.org/article/info:doi/10.1371/journal.pone.0080031

    ReplyDelete
    Replies
    1. I think I saw it but I had not bookmarked it in any case, so thanks for calling my attention to it. My first impression (but I have to read deeper) is that the phylogeny is the same in essence and that it shows two main branches: L0d and L0a'b'f'k. The former is clearly Southern African and the latter clearly East African. There is one branch (L0k) which seems to have starred some sort of migration to Southern Africa, however L0k is not so much concentrated in Southern Africa proper but rather to the North, with center in Zambia, Katanga and such (http://forwhattheywereweare.blogspot.com/2013/03/khoe-san-matrilineages-and-prehistory.html). Also, Barbieri 2013 (http://www.cell.com/AJHG/abstract/S0002-9297%2812%2900644-1) said that L0k in Southern Africa was mostly Bantu, not Khoisan:

      "Our major finding is the definition of two ancient sublineages of L0k (L0k1b and L0k2) that are present almost exclusively in Bantu-speaking populations from Zambia; the presence of such relic haplogroups in Bantu speakers is most probably due to contact with ancestral pre-Bantu populations that harbored different lineages than those found in extant Khoisan".

      So it's likely that L0k has a northernly origin and that it was brought to the South by Bantu peoples in essence. At least from previous studies.

      In this paper, there is one L0k sublineage only found in non-Bantu and another which is ambiguous, while the third one is Bantu-only, while the map shows frequency peaks in Zimbabwe and NE South Africa, areas of greatest Bantu consolidated expansion.

      In Pickrell 2012 (http://forwhattheywereweare.blogspot.com/2012/07/khoisan-autosomal-genetics.html), which deals with Khoisan autosomal genetics, the Darama appear Bantu-like (very strongly so). Other populations of the same linguistic group (Khoe-Kwadi) also appear variably influenced by Bantu or generally non-Khoisan genetics. Not sure how this relates with mtDNA but it's clear that we should make some distinction between the various Khoisan populations

      I don't have much time right now to make the appropriate search and research, but I'm pretty sure that there are studies that claim that L0k is related to a pastoralist pre-Bantu flow from East Africa, which had only limited genetic impact.

      I'd need more time to gather all the pieces into a comprehensive material but my impression right now is that L0 remains a dual Eastern and Southern African lineage and, that considering, the overall human original mtDNA genetics, an Eastern African origin for L0 remains the most likely scenario. Also using my usual method of counting (coding region) transitions from the root, I got that L0a'b'f'k should have coalesced long before L0d, which would be contemporary only of L0a'b'f. See: http://forwhattheywereweare.blogspot.com/2013/04/synthesis-of-spanish-language-series-on.html

      This is not wholly inconsistent with this paper's estimates, although I'm pretty sure I'd argue for some older dates in any case. I would say that the expansion of L0d, based on the archaeological record, should correspond to either c. 165 Ka BP (first MSA) or c. 130 Ka BP (clear increase in the density of findings). In any case the presence of H. sapiens and MSA is clearly older in Eastern Africa.

      But well, everything is opinable.

      Delete