Tuesday, June 26, 2012

Worldwide Distribution of Dodecad K10a Components

Numerous ADMIXTURE runs have been completed by the Dodecad Ancestry Project since its' inception approximately two years ago. The status of certain components remained tenuous despite subsequent runs, whilst others provided fairly stable values for the bulk of the project's participants.

With the completion of the latest K10a run, I have composed a series of geographically accurate frequency maps with the intention of effectively presenting the trends that can be seen through the raw data.


Method

Data; values from over 130 groups obtained through the Dodecad K10a Spreadsheet. Only groups with at least 5 participants considered. Composites of populations were taken where appropriate and denoted with _cmp. Labels shown otherwise identical to source. The O_Italian_D group was excluded because no information on their origins were found online. 

Mapping; Dodecad participant populations allocated to national capitals. Exact location of reference populations obtained where possible (see Citations) however some allowances were made regarding those accompanied by scant information. Refer to the Data Sink for the population list, coordinates and commentary made during mapping process. No numerical data, aside from those shown for certain populations, was shown to minimise clutter and to remain faithful to the intention of this entry.

Population depiction; I deemed it necessary to separately consider the genetic structure of Jewish, Indian and expatriate/New World populations and exclude them from the rest of Europe, Asia or Africa. Including Jewish minorities with their gentile compatriots would render the maps uninformative. The complexity of India's demographics, particularly because of the caste system, makes frequency maps an improper choice for revealing inter-group genetic differences. 


Results






















Acknowledgement

The raw values used in this investigation are attributed to Dienekes Pontikos, author of the Dodecad Ancestry Project.


Addenum I [04/07/2012]: Inclusion of All Components Colourised map, shown below:




Citations
http://www.uvm.edu/~rsingle/stat295/F05/papers/Cavalli-Sforza-NRG-2005_Ceph-HGDP-CDP.pdf
http://www.1000genomes.org/about 

6 comments:

  1. This comment has been removed by the author.

    ReplyDelete
  2. The absence of Kazakhs in Dienekes' analysis and hence in your maps significantly affects the distribution of the components in your maps. If Kazakhs were included, the Mongoloid components would show up much more prominent in Central Asia and thus the componental distributions of Central Asia would appear much closer to the truth.

    As for the O_Italian_D population, it stands for Other Italian. Thus it consists of all ethnic Italian Dodecad participants who don't belong to a regional Italian Dodecad population (e.g., C_Italian_D). So an Italian Dodecad participant who is, say, a Central Italian and South Italian mix would be assigned to the O_Italian_D population.

    Dienekes explained the meaning of O_Italian_D in this Dodecad thread (in the Comments section):

    http://dodecad.blogspot.com/2011/06/design-of-dodecad-v3.html

    ReplyDelete
    Replies
    1. Thank you for your post, as well as confirming the nature of the O_Italian_D group.

      "The absence of Kazakhs in Dienekes' analysis and hence in your maps significantly affects the distribution of the components in your maps. If Kazakhs were included, the Mongoloid components would show up much more prominent in Central Asia and thus the componental distributions of Central Asia would appear much closer to the truth."

      If the Harappa and Eurogenes Kazakh samples are anything to go by, I do not foresee their inclusion significantly altering the overall trends observed on these maps.

      The Kazakhs sit in the middle of an adequately sampled bottom circumference (Caucasus, South-Central Asia, East-Central Asia), meaning their inclusion would only "flesh out" the centre of that region only.

      This is, clearly, a limitation from the Dodecad Ancestry Project's current membership.

      Delete
    2. If the Harappa and Eurogenes Kazakh samples are anything to go by, I do not foresee their inclusion significantly altering the overall trends observed on these maps.

      The Kazakhs sit in the middle of an adequately sampled bottom circumference (Caucasus, South-Central Asia, East-Central Asia), meaning their inclusion would only "flesh out" the centre of that region only.


      Indigenous Siberian populations, too, are fairly well represented. So, you are correct in proposing that the inclusion of Kazakhs would only flesh out the center of that region. I used the expression "significanly affects" due to the largeness of Central Asia and the largeness of its part that Kazakhs inhabit.

      Delete
  3. The small SW Asian/Mediterranean components east(Central and South Central Asia) of Iran are minimal. Is this because those areas never received what is a second Neolithic wave that Iran did probably corresponding to R1b/J1? But they still exist. Is this admixture from various ethncities settling the area during Persian empire/Silk Road times? I have heard about Armenian/Assyrian/Jewish settlement during various empires. Although that doesn't sound like a plausible theory to me I have read it. http://www.agbu.org/publications/article.asp?A_ID=52 I originally though of North Caucasian/Russian/Tatar/Chuvash admixture inflating Northern European in Central Asia but the lack of I,N and specific G clades speaks against that. I do recall soem Tajik R1b but I would guess that is of Western Indo-Iranian origins and not some sort of admixture from non Indo-Iranian groups. The Western Euro component is interesting. Did Eastern Europe maybe have Western European admixture at the time the

    I'm also curious as to what the ancient breakdown of pre Indo-Iranian Iran/South Central Asia in addition to what the breakdown of components would have been in Iranian Central Asia. Especially areas like Tukrmenistan (due to what seems to be seemingly higher Iranian plateau ancestry as seen through the SW Asian/Med components present compared to Afghanistan or Tajikistan) and Kazakhstan due for it encompassing most of Andronovo territory and having Srubnaya settlements. Afghanistan/Tajikistan/Kyrgzstan for their Tocharian settlements (and Iranian plateau ancestry for Tajiks) would be interesting as well. There was some Scythian back migration due to reasons people are not sure of but I doubt that was anything massive and probably didnt change either Europe or Central Asia too much. Too bad we don't have aDNA or even any idea about what craniometric/pigmentation phenotypes were present.

    On the other hand Atlantic-Baltic (although I am still confused on what the Western Euro part of this component represents as western admixture in eastern Europe is relatively recent and accompanied by Ydna I1/R1b and that can't be the case for Andronovo)is weaker in Siberia than Central Asia so I don't think much if any of this component among Turks comes up from Admixture with proto Turks or Turkic admixed uralic speakers. I have heard of the theory of portions of Central Asia(particularly Northern/Western Kazakhstan but Northern kazakhstan seems to be constantly changing since Scythian times in its majority population in history as we know the Russians are most numerous there today whereas-this makes sense since Northern kazakhstan is a steepe) being Uralic in ancient times even up to Andronovo having some Uralic speakers in its fold. No idea if this is true. I haven't seen any evidence for Uralic types present anywhere except North of Kazakhstan/South Russia in ancient times.

    It does seem clear that South Asia and Central Asia had different Neolithics as evidenced by Uzbeks having the Mediterranian component. Also mit seems like the incoming Proto Indo-Iranians were 100% Northern European with no Mediterrranian admixture? Since NW south Asian groups lack Med? I doubt West Asian admixture was present because the Neolithic farmers IE groups mixed with were Yamnaya farmers. Can't see much of a ydna/tmdna contribution from there. But they apparently spread lactose persistence to steepe tribes. And apparenlty impacted the phenotypes:http://dienekes.blogspot.com/2005/04/anthropology-of-sredny-stog-and.html. But hard to see any tripoyle ydna unless things like ydna I in Central Asia just ended up dying out because they were only present at small frequencies inA ndronovo g roups. I don't think that appplies to N1c and Uralic groups who lived in the forsts north of Andronovo I believe and didn't occupy Kazakstan or the Sintash Petrovka culture land in modern day Russia.

    ReplyDelete
  4. That band of faint Southeast Asian along the Indian Ocean coast of Africa caught my eye, but in reviewing the spreadsheet I see it's artifactual rather than an indication of Austronesian touchdown.

    ReplyDelete