PDA

View Full Version : Mota ancestry in Northern Ethiopia & Eritrea



drobbah
03-07-2021, 02:44 AM
Why are there some Habeshas that don't seem to score any Mota on G25 while there are others who score significant percentages? The Cushitic Agaw speakers seem to have it at higher levels too despite being generally similar in terms of their MENA/SSA ratio to their Habesha neighbours.What's the historical reasoning behind this? I thought since Horners have been in that region for thousands of years that Mota levels would of been homogeneous like the Agaw (except the outlier) by now but even between the Tigray or Eritreans there seems to be a gap.Amharas are understandable as they were affected to some degree by the Oromos and were quite open to assimilation of others in their region.


Habeshas

https://i.imgur.com/Eo0adQ4.jpg
https://i.imgur.com/QZtgds1.jpg
https://i.imgur.com/eZ3bdYA.jpg



Cushitic Agaw speakers including the Jews

https://i.imgur.com/lmc4EF2.jpg

ThaYamamoto
03-07-2021, 07:37 PM
Because

1. g25 is a PCA so the ancestral contributions are bound by its dimensions
2. the fits are atrocious because of a lack of true ancestral sources/proxies so you
3. run into statistical artefact(s) like the results of your models
4. mota is barely differentiated from dinka on on dimensions after PC2
5. deep ancestry for africans via g25 is a faux-pas particularly without the ancients required

https://i.ibb.co/KG2C1dw/motasudansomalia-PCA.png (https://ibb.co/QcPs5Ld)

https://i.ibb.co/JvMyPQL/motasudansomlia2.png (https://ibb.co/dPzLyg9)

drobbah
03-07-2021, 07:46 PM
I ran the other Horners like the Afars (very low Mota),Somalis (mostly low Mota) and the Oromos (high Mota).So clearly the model works although the Mota scores shouldn't be taken at face value obviously but it clearly is good at showing a certain pattern.The Mota scores for me in this model is quite close to what I get on qpadm also.

It seems clear there are individuals within the Habesha who don't have much Mota while there are others rich in Mota ancestry.Hopefully someone can get high coverage Horner samples and run them on qpadm

Horner
03-07-2021, 08:36 PM
My first thought is also that there may be some noise in the G25 model. How does this Mota reference used in your your model compare to Omotic cluster used in PuntDNAL's K15 admixture run? I assume the Puntdnal run uses modern populations that may have other components. In the admixture run we see a NE to SW cline amongst people from the Horn (Afars/Somali showing the lowest and Wolayta the highest), but the variation between Tigrinya speakers is negligible (~1%).

Also, I am not sure where your Eritrean 2 & 3 samples are from, but if they are part lowland Tigre, Saho or Hidareb it might explain the lower Omotic signal.

drobbah
03-07-2021, 08:38 PM
My first thought is also that there may be some noise in the G25 model. How does this Mota reference used in your your model compare to Omotic cluster used in PuntDNAL's K15 admixture run? I assume the Puntdnal run uses modern populations that may have other components. In the admixture run we see a NE to SW cline amongst people from the Horn (Afars/Somali showing the lowest and Wolayta the highest), but the variation between Tigrinya speakers is negligible (~1%).

Also, I am not sure where your Eritrean 2 & 3 samples are from, but if they are part lowland Tigre, Saho or Hidareb it might explain the lower Omotic signal.
I think the high Mota individual is Tigrayan and the other two are lowlander Tigre

drobbah
03-07-2021, 08:56 PM
For comparison purposes with other Horners:

Somalis & Rendille
https://i.imgur.com/JpNhAQp.jpg

Afars
https://i.imgur.com/Hsze7gY.jpg

Oromos & Wolayta
https://i.imgur.com/oq7PkB9.jpg

predominantly non-Cushitic descended Ethiopians
https://i.imgur.com/I53v0gc.jpg
https://i.imgur.com/3fZSuJu.jpg
https://i.imgur.com/q22yvcu.jpg