Tuesday, December 18, 2012

Ramesses III belonged to YDNA haplogroup E1b1a


According to a study published yesterday, Revisiting the harem conspiracy and death of Ramesses III: anthropological, forensic, radiological, and genetic study, Y- STR data places his YDNA haplogroup in E1b1a using Whit Athey's Haplogroup predictor: 

"Genetic kinship analyses revealed identical haplotypes in both mummies (table 1); using the Whit Athey’s haplogroup predictor, we determined the Y chromosomal haplogroup E1b1a. "


his DYS repeats are listed as follows:

DYS 19 19
DYS 385a,b 20
DYS 389I 13
DYS 389II 33
DYS 390 21
DYS 391 8
DYS 392 17
DYS 393 8
DYS 437 14
DYS 438 10
DYS 448 20
DYS 456 13
YGATAH4 13


Plugging these numbers in Whit Athey's predictor does indeed indicate that his haplogroup is E1b1a with 99.1% probability using equal priors. The decisive DYS, to judge between E1b1a and E1b1b, is DYS 390, with the exclusion of DYS 390, his haplotype belongs to 83.7 % E1b1b and 15.8% E1b1a, however, it is well known that DYS 390 = 21 is a high probability signature for West/Central/Southern Africa, i.e. where E1b1a dominates (see below).


UPDATE : Upon receiving an e-mail stating that the haplotype could still belong to E1b1b, on the basis of a haplotype from Chad present in the FTDNA database that has DYS 390 = 21, I further looked into it, the presence of such a haplotype would not necessarily refute what the authors of this study are claiming, because if one enters the repeats for those Chad E1b1b haplotypes (but only for the same DYS#'s that are included this study) into the predictor, i.e. :


DYS 19 13
DYS385a,b 15
DYS 389I 12
DYS 389II 29
DYS 390 21
DYS 391 9
DYS 392 11
DYS 393 13
DYS 437 14
DYS 438 10
DYS 448 21
DYS 456 15
YGATAH4 11




one would still get an assignment to haplogroup E1b1b with 88.6% probability and only a 5.6% probability that the haplotype may belong to E1b1a, therefore, I highly doubt that this pharaoh's haplotype, if extracted correctly and with out contamination, would be anything but E1b1a, absent an SNP test however, one can never be 100% sure.

6 comments:

  1. Please clarify your last paragraph saying "anything but e1b1a"

    You stated 88.6% likelihood of being e1b1b, so you are saying that you doubt it would be anything but e1b1b?

    ~ Mike Dammann
    http://www.rhesusnegative.net

    ReplyDelete
    Replies
    1. The communication I received implied that the predictor was simply just confusing Ramesses' E1b1b haplotype for an E1b1a haplotype due to DYS 390 being equal to 21 repeats, since in-fact, there are (rare as they maybe) cases of E1b1b haplotypes that also have DYS 390 = 21, the email pointed to the Chadian FTDNA haplotype as one exemplary sample.

      I thought that was an interesting supposition, however, it was wrong, why? because when you enter those SNP tested Chadian Y-STR in the predictor, they are not incorrectly predicted as E1b1a, which would be the case we would expect if the Pharoah's haplotype were being genuinely confused, rather the Chadian haplotype was correctly being predicted as E1b1b. Is that any clearer ?

      Delete
  2. Yes, thank you. So in other words: We have narrowed things down to two good guesses one of which would be certain.

    Another question: Are there other indicators you might be aware of which could lead me into the right direction?
    I would like for you to confirm whether you have credible data on the following:

    1) Claims that Ramses II was HLA-B27 positive.
    2) That the cde frequency amongst mummie ramains tested for blood type was high (the socalled "rh negative gene")

    I am getting claims thrown at me that Ramses II was supposedly B negative and would like for you to let me know what your take on this is.

    Thank you in advance for your help.

    ~ Mike

    ReplyDelete
  3. Marianne Luban put the STRs of Ramses III into popaffiliator program and they presented as sub-Saharan, as well.

    http://thetimetravelerreststop.blogspot.fr/2012/12/here-is-what-came-out-when-i-put-dna-of.html


    To further check the validity of popAffiliator (http://cracs.fc.up.pt/~nf/popaffiliator2/) using the same 9 STRs, I ran the STRs of a few samples from DNATribes website (http://www.dnatribes.com/sampleresults.html). I also used popAffiliator 2 as it is a liitle bit more accurate that popAffiliator 1.

    As we can see below, ALL of them were assigned to their correct Continental groups (Sub-Saharan Africa, Eurasia, Asia).

    Just note these are NOT admixture proportions but just probabilities. So it seems that even 8 markers are enough to assign the correct continental groups with Sub-Saharan individuals always getting a Sub-Saharan probability > 90%.

    9 STRS used(CSF1PO,D2S1338, D3S1358,D13S317, D16S539, D18S51, D21S11, FGA, D7S820)

    Ramses III:

    Sub-Saharan Africa 96.2%
    Eurasia 2%
    Asia 1.8%

    (D3S1358 was defaulted as mentionned by Marianne but using any value for D3S1358, Ramses III always gets a probablity > 90% for sub-Saharan Africa.)

    1) Sub-saharan individuals

    African America:
    Sub-Saharan Africa 99.2%
    Eurasia 0.6%
    Asia 0.3%

    Uganda:
    Sub-Saharan Africa 93%
    Eurasia 4.3%
    Asia 2.7%

    2) Eurasian individuals

    South Italian:
    Eurasia 59.2%
    Asia 28.8%
    Sub-Saharan Africa 12%

    Berber:
    Eurasia 47.4%
    Sub-Saharan Africa 37%
    Asia 15.6%

    Saudi Arabian:
    Eurasia 45.6%
    Sub-Saharan Africa 42.7%
    Asia 11.6%

    Scottish:
    Eurasia 63.4%
    Sub-Saharan Africa 31%
    Asia 5.6%

    Turkish:
    Eurasia 62.6%
    Sub-Saharan Africa 19.6%
    Asia 17.8%

    3) Asian individuals

    Korean:
    Asia 77%
    Eurasia 18.6%
    Sub-Saharan Africa 4.4%

    ReplyDelete
    Replies
    1. Interesting, for the sake of precision, the populations used to train the machine learning model to identify Subsahran Africans are 195 Namibians, 404 Somalis, 225 Tanzanians and 47 unspecified SSA.

      The publication for popaffiliator, Pereira et.al can be found here https://oda.hio.no/jspui/bitstream/10642/472/2/529937post.pdf

      Another point is that there are 3 datasets that the publication specifies, one is the 3R, which encompasses West Asia, SSA and East Asia, apparently the dataset that your friend used above, however there are 2 more, the 5R which encompasses “Asia, Eurasia, Sub-Saharan Africa, North Africa and Near East”, and the 7R, which encompasses all the 7 regions in the database. I suspect using the other 2 datasets may yield some different results, note that the North African training populations includes, among others, Egyptians from Siwa and Southern Egypt, and possibly berbers from Morroco as well.

      Delete
  4. DnaTribes just released a new study which seems to confirm "evidence of Sub-Saharan African ancestry (possibly among several ancestral components) for pharaonic families of ancient Egypt".

    "Digest for February 2013: Ramesses III and African Ancestry in the 20th Dynasty of New Kingdom Egypt" : http://www.dnatribes.com/dnatribes-digest-2013-02-01.pdf

    ReplyDelete