Predynastic Egyptian genome's ancestry broken down

[Brandon S. Pilcher]

Courtesy of ES's own Elmaestro (aka Stro or Revoiye):

Pre-dynastic Egyptian DNA: A Sneak Peak into North East Africa’s Distant Past

This analysis points to this pre-dynastic sample being similar to Epipaleolithic and early Neolithic Moroccans in that they posses a genetic profile vaguely related to both West Eurasians and Sub-Saharan Africans. More so, they typically fail to yield statistically significant results from two way admixture models of any modern Sub-Saharan and ancient West Eurasian source and the percentages of each cohort vary depending on the comparative samples or method of analysis.⁠ Though these North African samples can at times be used interchangeably as an “autochthonous” North African proxy, the two regions (North-West and North-East Africa ) seem to prefer different ancestral proxies themselves. For instance, while North-West Africans like Taforalt can be modeled most significantly with Natufians, AM14590 (the sample ID) of Naqada’s west Eurasian related DNA is most similar to Neolithic Iranians. In fact, it seems that the individual unearthed from one of the Belt Caves, Hotu, shares the most similarity with the “Non-African” portion of this pre-dynastic sample. Though, it could be the case that Hotu’s relatively low SNP count yielded high P values and ancestry estimates due to ascertainment bias, the overwhelming preference of this sample to model the bulk of the pre-dynastic samples ancestry should still be noted. When using the Natufian and Yoruba to model Naqada I the proportions are more or less identical to that of Epipaleolithic and Early Neolithic Moroccans though the p-values are close to null.

Keep in mind that those Neolithic Iranian samples have the most ancestry from a ghost population called "Basal Eurasian", which is related to OOA populations but lacks Neanderthal ancestry and could very well be of geographic Northeast African origin. So this one predynastic sample might be even more African in ancestry than the statement "vaguely related to both West Eurasians and sub-Saharan Africans" might imply.

Also, a bar graph showing how much predynastic ancestry later Egyptian (and Nubian) samples have:


And finally some phenotype calls for the samples, including skin color (an alternate version I annotated myself to show sample provenance is here). Note that most of these AE samples are predicted to have had "dark to black" skin.

What do you guys think of this?

[djehuti]

^ If we use Naqada_Gebelein marker as the baseline for what is 'Egyptian' the results are indeed interesting. Note how except for one, the Abusir samples shown have NO Naqada marker and the one that does, has it at low frequency. The 5 samples with the highest frequency of Naqada marker are foremost the Late Period Unknown sample that also has a little Dinka-like marker, followed by the Middle Kingdom Rifeh NK sample with slightly higher Dinka marker, followed by Middle Kingdom Rifeh NA sample with Levant Neolithic marker that is higher, and then the Third Intermediate Period Unknown, and of course the Kadruka sample.

Compare with the older Revoiye admixture run on Kadruka.



Also the 2017 paper on modern populations of Northeast Africa.

[Brandon S. Pilcher]

Is it just me, or is the proportion of Dinka-related ancestry in most of Elmaestro's AE sampples not that much higher than the pooled modern Egyptian samples? Most of them have a least a little Dinka. Since a lot of these new samples come from Upper Egyptian sites IIRC, it would be interesting to compare them to modern Upper Egyptians instead of just modern Egyptians in general. I know Swenet has suggested that dark-skinned modern Upper Egyptians might have enough elevated SSA relative to the ancients to affect their phenotype somehow.

[Brandon S. Pilcher]

BTW, I want to see how much of this predynastic Egyptian-like ancestry exists in ancient Levantine remains. How mixed with Egyptians would Bronze Age peoples of the region like the Canaanites and Hebrews would have been?

[maverick5]

In light of these new results if Naqada is indeed 60% Eurasian and 40% Sub-Saharan related isn't that less African compared to Taforalt/Iberomaurusians whom were modelled at 45% ANA and 55% West Eurasian? Why does Taforalt then cluster so far from Pre-Dynastic Egyptians and every other Black African for that matter if they've slightly more African related ancestry than this Naqada sample?

[Shadow]

Because Taforalt has different Eurasian ancestry from the Naqada sample. Taforalt can be modeled from a source best represented by Dzudzuana, which was more excessively Western Eurasia, WHG. The Naqada sample's Eurasian ancestry can be best modeled by Iranian Hunter Gatherers, who has the most Deep (Basal Eurasian) ancestry. This difference in Eurasian ancestry can create the discrepancies needed to make the IBMs divergent from predynastic Egyptians, as the IBMs were robust and Western Eurasian ancestry or in this case WHGs were shown to be correlated with robusticity. We can only hypothesized how the Basal Eurasians could've looked. Bare in mind that part of this craniometric divergence could be due to the two groups African ancestry, whether from the IBMs' Aterian ancestry, YAP ancestry, or both. The ancestors of the predynastic Egyptians were also roughly robust too, but lost it through advent of agriculture revolution. Also, there's no way of telling if the ratio in African vs. Eurasian ancestry from the Naqada sample is precise to say the least. Neolithic and Mesolithic Iranians were shown to have discernable SSA ancestry in several models, so some of the Naqada sample's SSA ancestry could be absorbed by Iranian HG ancestry, similar to how the IBMs' SSA ancestry was absorbed when the Natufians were used as a proxy for their Eurasian ancestry.

[Shadow]

Btw, another thing each of you should take into account is the time and space of this sample. The sample comes from Gebelein, a site that received bidirectional geneflow from Northern Egypt. This is demonstrated by the craniometric divergence of OK/Int. Gebelein samples from other predynastic Egyptians such as the Badarians, who cluster knit closely with a later Nubian series (Aksha is a site that had greater Equatorial influence btw).



As you all should know, cosmopolitan centers of Southern Egypt became more diversified from the Naqada period and onwards, which likely explains the ratios of African vs. Eurasian ancestry, that is if the models are precise. Also likely explains why Neolithic & Mesolithic Iranians are the best proxies for the predynastic sample's Eurasian ancestry in comparison to just Dzudzuana. It's not that the Naqada sample derives all of their Eurasian ancestry from Neolithic/Mesolithic Iranians, only that predynastic Egyptians received geneflow from a Iranian source, which then fused in with pre-existing Eurasian ancestry they would've had. Kind of similar to how the best source of ancestry for modern Ethiopians were from the Minoans, despite there being no direct relationship between the two groups. Ethiopians just received geneflow that mixed in with the Eurasian ancestry they already had, creating the discrepancy. Btw, Neolithic Iranian ancestry was already present in the Levant during the Chalcolithic period, so this theory bares out in my opinion.

[maverick5]

Jul 29, 2024 12:56:43 GMT -5 Shadow said:

Because Taforalt has different Eurasian ancestry from the Naqada sample. Taforalt can be modeled from a source best represented by Dzudzuana, which was more excessively Western Eurasia, WHG. The Naqada sample's Eurasian ancestry can be best modeled by Iranian Hunter Gatherers, who has the most Deep (Basal Eurasian) ancestry. This difference in Eurasian ancestry can create the discrepancies needed to make the IBMs divergent from predynastic Egyptians, as the IBMs were robust and Western Eurasian ancestry or in this case WHGs were shown to be correlated with robusticity. We can only hypothesized how the Basal Eurasians could've looked. Bare in mind that part of this craniometric divergence could be due to the two groups African ancestry, whether from the IBMs' Aterian ancestry, YAP ancestry, or both. The ancestors of the predynastic Egyptians were also roughly robust too, but lost it through advent of agriculture revolution. Also, there's no way of telling if the ratio in African vs. Eurasian ancestry from the Naqada sample is precise to say the least. Neolithic and Mesolithic Iranians were shown to have discernable SSA ancestry in several models, so some of the Naqada sample's SSA ancestry could be absorbed by Iranian HG ancestry, similar to how the IBMs' SSA ancestry was absorbed when the Natufians were used as a proxy for their Eurasian ancestry.

I thought the Natufians had the most  Basal Eurasian ancestry?



This here is saying the Natufians had 50% Basal Eurasians on average.

[Shadow]

Jul 29, 2024 14:43:27 GMT -5 maverick5 said:

Jul 29, 2024 12:56:43 GMT -5 Shadow said:

Because Taforalt has different Eurasian ancestry from the Naqada sample. Taforalt can be modeled from a source best represented by Dzudzuana, which was more excessively Western Eurasia, WHG. The Naqada sample's Eurasian ancestry can be best modeled by Iranian Hunter Gatherers, who has the most Deep (Basal Eurasian) ancestry. This difference in Eurasian ancestry can create the discrepancies needed to make the IBMs divergent from predynastic Egyptians, as the IBMs were robust and Western Eurasian ancestry or in this case WHGs were shown to be correlated with robusticity. We can only hypothesized how the Basal Eurasians could've looked. Bare in mind that part of this craniometric divergence could be due to the two groups African ancestry, whether from the IBMs' Aterian ancestry, YAP ancestry, or both. The ancestors of the predynastic Egyptians were also roughly robust too, but lost it through advent of agriculture revolution. Also, there's no way of telling if the ratio in African vs. Eurasian ancestry from the Naqada sample is precise to say the least. Neolithic and Mesolithic Iranians were shown to have discernable SSA ancestry in several models, so some of the Naqada sample's SSA ancestry could be absorbed by Iranian HG ancestry, similar to how the IBMs' SSA ancestry was absorbed when the Natufians were used as a proxy for their Eurasian ancestry.

I thought the Natufians had the most  Basal Eurasian ancestry?



This here is saying the Natufians had 50% Basal Eurasians on average.

Well, you thought wrong. That title goes to Neolithic/Mesolithic Iranians:



Neolithic Iranians were also estimated of having around ~50% Basal Eurasian ancestry just like the Natufians. However, Mesolithic Iranians were estimated as having roughly ~60% Basal Eurasian ancestry.

We obtain the highest estimates in the earliest populations from both Iran (66±13% in the likely Mesolithic sample, 48±6% in Neolithic samples), and the Levant (44±8% in Epipaleolithic Natufians) (Fig. 2), showing that Basal Eurasian ancestry was widespread across the ancient Near East.

Source: www.biorxiv.org/content/10.1101/059311v1.full

The chart I posted was technically measuring "Deep ancestry", as opposed to Basal Eurasian ancestry itself, but you get the point. Also, bare in mind that these are old models for these groups ancestry and that they have been shown from recent models to have some discernable SSA ancestry. This table from Revoiye for example shows a Neolithic Iranian sample to be around 4% SSA:

[djehuti]

Jul 23, 2024 16:16:17 GMT -5 Brandon S. Pilcher said:

Is it just me, or is the proportion of Dinka-related ancestry in most of Elmaestro's AE samples not that much higher than the pooled modern Egyptian samples? Most of them have a least a little Dinka. Since a lot of these new samples come from Upper Egyptian sites IIRC, it would be interesting to compare them to modern Upper Egyptians instead of just modern Egyptians in general. I know Swenet has suggested that dark-skinned modern Upper Egyptians might have enough elevated SSA relative to the ancients to affect their phenotype somehow.

I remember Swenet suggesting that such Dinka-like ancestry may very well be part of the basic general Nile Valley ancestry so that it wouldn't be unusual to see little amounts here and there among predynastic Egyptians. Though as Swenet has also pointed out the many anthropologists who suggested increasing amounts of Sub-Saharan ancestry from at least the Middle Kingdom to later times.

Jul 27, 2024 21:27:50 GMT -5 Brandon S. Pilcher said:

BTW, I want to see how much of this predynastic Egyptian-like ancestry exists in ancient Levantine remains. How mixed with Egyptians would Bronze Age peoples of the region like the Canaanites and Hebrews would have been?

Especially since we are able to differentiate this Egyptian ancestry from the Neolithic Levant/Natufian? Although Swenet also suggests that the late period of Natufians going into the Neolithic show crania remains similar to Egyptians than older period Natufians.

[djehuti]

Jul 29, 2024 12:56:43 GMT -5 Shadow said:

Because Taforalt has different Eurasian ancestry from the Naqada sample. Taforalt can be modeled from a source best represented by Dzudzuana, which was more excessively Western Eurasia, WHG. The Naqada sample's Eurasian ancestry can be best modeled by Iranian Hunter Gatherers, who has the most Deep (Basal Eurasian) ancestry. This difference in Eurasian ancestry can create the discrepancies needed to make the IBMs divergent from predynastic Egyptians, as the IBMs were robust and Western Eurasian ancestry or in this case WHGs were shown to be correlated with robusticity. We can only hypothesized how the Basal Eurasians could've looked. Bare in mind that part of this craniometric divergence could be due to the two groups African ancestry, whether from the IBMs' Aterian ancestry, YAP ancestry, or both. The ancestors of the predynastic Egyptians were also roughly robust too, but lost it through advent of agriculture revolution. Also, there's no way of telling if the ratio in African vs. Eurasian ancestry from the Naqada sample is precise to say the least. Neolithic and Mesolithic Iranians were shown to have discernable SSA ancestry in several models, so some of the Naqada sample's SSA ancestry could be absorbed by Iranian HG ancestry, similar to how the IBMs' SSA ancestry was absorbed when the Natufians were used as a proxy for their Eurasian ancestry.

I'm curious. Do you Shadow identify 3 possibly related lines of ancestry among IBMs, Predynastic Egyptians, and Basal Eurasians? For example, I tend to go with the hypothesis that all three share E-M35 paternal clade with IBM having E-M81, Egyptians having E-M78, and Basal Eurasians having E-M34. I also identify the last group to carry the Arabian-Indian form of HbS.

Jul 29, 2024 13:40:49 GMT -5 Shadow said:

Btw, another thing each of you should take into account is the time and space of this sample. The sample comes from Gebelein, a site that received bidirectional geneflow from Northern Egypt. This is demonstrated by the craniometric divergence of OK/Int. Gebelein samples from other predynastic Egyptians such as the Badarians, who cluster knit closely with a later Nubian series (Aksha is a site that had greater Equatorial influence btw).



As you all should know, cosmopolitan centers of Southern Egypt became more diversified from the Naqada period and onwards, which likely explains the ratios of African vs. Eurasian ancestry, that is if the models are precise. Also likely explains why Neolithic & Mesolithic Iranians are the best proxies for the predynastic sample's Eurasian ancestry in comparison to just Dzudzuana. It's not that the Naqada sample derives all of their Eurasian ancestry from Neolithic/Mesolithic Iranians, only that predynastic Egyptians received geneflow from a Iranian source, which then fused in with pre-existing Eurasian ancestry they would've had. Kind of similar to how the best source of ancestry for modern Ethiopians were from the Minoans, despite there being no direct relationship between the two groups. Ethiopians just received geneflow that mixed in with the Eurasian ancestry they already had, creating the discrepancy. Btw, Neolithic Iranian ancestry was already present in the Levant during the Chalcolithic period, so this theory bares out in my opinion.

Yes, I'm reminded of the various studies confirming what Batrawi found-- that once Egypt was united there was more gene flow from both north and south with northern types appearing in the major cities of the south. I am curious about the Badarians. All the cranial studies I've read both metric and nonmetric make the Badarians out to be the centroid population in all predynastic and dynastic studies. And we all know that the main difference between Badarians and Naqada is that the latter is taller and more orthognathous. I just wonder what the genetic differentiation is between Badarian and Naqada. I also wonder about the little Hadza-like markers we see in Dzudzuana according to the Loosdrecht paper.

[zarahan]

Djehuti says:
Yes, I'm reminded of the various studies confirming what Batrawi found-- that once Egypt was united there was more gene flow from both north and south with northern types appearing in the major cities of the south. I am curious about the Badarians. All the cranial studies I've read both metric and nonmetric make the Badarians out to be the centroid population in all predynastic and dynastic studies. And we all know that the main difference between Badarians and Naqada is that the latter is taller and more orthognathous. I just wonder what the genetic differentiation is between Badarian and Naqada. I also wonder about the little Hadza-like markers we see in Dzudzuana according to the Loosdrecht paper.

Good question. I think as ADNA studies expand a more detailed profile will emerge.
I am sure neither Badara or Naqada were totally isolated from other population
movements or influences in the Nile Valley, on into Sinai, given centuries of time.

[zarahan]

Jul 22, 2024 0:05:45 GMT -5 Brandon S. Pilcher said:

Courtesy of ES's own Elmaestro (aka Stro or Revoiye):

Pre-dynastic Egyptian DNA: A Sneak Peak into North East Africa’s Distant Past

This analysis points to this pre-dynastic sample being similar to Epipaleolithic and early Neolithic Moroccans in that they posses a genetic profile vaguely related to both West Eurasians and Sub-Saharan Africans. More so, they typically fail to yield statistically significant results from two way admixture models of any modern Sub-Saharan and ancient West Eurasian source and the percentages of each cohort vary depending on the comparative samples or method of analysis.⁠ Though these North African samples can at times be used interchangeably as an “autochthonous” North African proxy, the two regions (North-West and North-East Africa ) seem to prefer different ancestral proxies themselves. For instance, while North-West Africans like Taforalt can be modeled most significantly with Natufians, AM14590 (the sample ID) of Naqada’s west Eurasian related DNA is most similar to Neolithic Iranians. In fact, it seems that the individual unearthed from one of the Belt Caves, Hotu, shares the most similarity with the “Non-African” portion of this pre-dynastic sample. Though, it could be the case that Hotu’s relatively low SNP count yielded high P values and ancestry estimates due to ascertainment bias, the overwhelming preference of this sample to model the bulk of the pre-dynastic samples ancestry should still be noted. When using the Natufian and Yoruba to model Naqada I the proportions are more or less identical to that of Epipaleolithic and Early Neolithic Moroccans though the p-values are close to null.

Keep in mind that those Neolithic Iranian samples have the most ancestry from a ghost population called "Basal Eurasian", which is related to OOA populations but lacks Neanderthal ancestry and could very well be of geographic Northeast African origin. So this one predynastic sample might be even more African in ancestry than the statement "vaguely related to both West Eurasians and sub-Saharan Africans" might imply.

Also, a bar graph showing how much predynastic ancestry later Egyptian (and Nubian) samples have:


And finally some phenotype calls for the samples, including skin color (an alternate version I annotated myself to show sample provenance is here). Note that most of these AE samples are predicted to have had "dark to black" skin.

What do you guys think of this?

I think El Maestro's analysis is intriguing as far as the skin color, and would
track with both other studies of skin color, as well as iconography and statuary
showing a range of dark skin colors. This would again debunk assorted Arabist, Nativist or
Eurocentric claims trying to make out that dark skin is somehow "foreign" to Egypt.

[zarahan]

Djehuti says:
Also, there's no way of telling if the ratio in African vs. Eurasian ancestry from the Naqada sample is precise to say the least. Neolithic and Mesolithic Iranians were shown to have discernable SSA ancestry in several models,

What's the study re the Neolithic and Mesolithic Iranians showing traces of SSA ancestry?

[djehuti]

^ Sorry for the late response but since I was posting in original ES. But now that it's down again...

One study I had in mind was the Loosdrecht et al. study showing both Iranian Neolithic and Anatolian Neolithic to possess Hadza ancestry.



The Schuenemann et al. study also shows the Iranian Neolithic to posses some pinkish looking markers.



There's also the skeletal evidence showing "negroid" traits not to mention the Arab-Indian HbS.