PDA

View Full Version : "Are All Africans Genetically North African?"



Brandon S. Pilcher
03-28-2021, 09:09 PM
This is a blog post by someone under the pseudonym "Stro" that aims to measure the amount of "Ancestral North African" ancestry in a variety of African populations, as well as a few from the ancient Near East.

Are All Africans Genetically North African? (https://revoiye.com/are-we-all-genetically-north-african/)

A chart from the above blog post listing estimated percentages of ANA ancestry:
https://i0.wp.com/revoiye.com/wp-content/uploads/2021/03/tableofanaestimatesfixed.png?w=962&ssl=1

I will admit, I am surprised to seeing ANA peak in a number of populations I would have assumed were typically "sub-Saharan" in affinity, such as South Sudanese Dinka and Shilluk. But the amount of ANA he finds in Taforalt and early Neolithic Moroccans is on par with that found in the Lazaridis papers.

By the way, he explains his methodology in this other post (https://revoiye.com/the-breakdown-are-we-all-genetically-north-african/).

Any thoughts on this?

drobbah
03-28-2021, 09:22 PM
Why is KEN_Kakapel at 87% and Dinka is at 74%? What's the main difference between these two populations?

theplayer
03-28-2021, 11:00 PM
This is great, pretty much in agreement with my own speculations about deep African population history.


I would personally have guessed West African having closer to ~70% ANA ancestry.
Going by uniparentals; basically 100% ydna E for most populations with mtdna ~5% U6 and ~35% L3 on average => 70% ANA-related ancestry(also in-line with the Shum-Laka paper).
Otherwise that ANA % table seems pretty accurate I'd say(of course AEA is probably confused for ANA, east african foragers(Nyarindi, Malawi Hora, Tanzanian Forager) likely don't have that much ANA even though I would guess it is some).

So basically what I've been thinking is: during the beginning of the (last) African Humid Period people from North Africa(ANA-like) would migrate all over the Sahara spreading "Niger-Congo" and "Nilo-Saharan"(NS languages don't need to be a real language family, or at least not all of them, though likely many had some contact early on at least) languages and the "Aquatic Civilization" HG culture. Pottery would soon be invented(my guess would be in the air mountains) rapidly spreading to Egypt/Sudan and Mali(Ounjougou) and the rest of the Sahara. Early forager pottery would even spread as far as Kenya early on, which could explain the haplogroup E presence in Mota-related populations(could be more ancient as well).

In the west "Niger-Congo" populations would form as admixed with "Basal Human"/"Ghost Modern"(what I'd call AWA, Ancient West Africans), in the east "Nilo-Saharan" with with some unknown(could be ACA-related) "deep" population + some AEA I'd imagine. Shum Laka would be an early mix of this Saharan(ANA) ancestry and in this case Ancient Central African(ACA) instead(with some AWA as well, see ShumLaka paper).
Then of course early Cushitic populations(I don't think afroasiatic comes from Natufian but from a population very similar to them in Egypt/Sudan) would likely be Natufian-like(Dzudzuana-like+Iberomaurasian-like) plus admixture from the new arrivals rich in ANA(Dinka being the usual proxy for this population).

The only problem would be that the sampled Iberomaurasians(Taforalt) had too much (female mediated, mtdna U6) Eurasian admixture. Though I could imagine a more "pure" ANA population living further south(taforalt is by the coast) perhaps the Atlas mountains? This less-eurasian Iberomaurasian population could still have like 5-10% eurasian ancestry(as there have been papers with Yoruba as about that amount Eurasian).

There are probably even worse problems with this hypothesis that I haven't considered that will render this pet theory of mine impossible though...

44059

Note: while I here drew the Eurasian in Iberomaurasian as coming from Iberia it could also have been from the Near-East.

AEA(which I associate with Mota, not Dinka) would have been haplogroups B and L4 while ANA would be E and L3 (so AEA splits from ANA+Eurasian which forms a (weak) clade somewhere around 70-90 kya, the ancestor of these population being what I call "Neo-African", associated with the early east-african LSA).

The following image was an attempt to somewhat take into account uniparentals into the phylogeny(A00, perhaps also A0(?) would be "GhostArchaic" related haplogroups)
44060

Michalis Moriopoulos
03-28-2021, 11:12 PM
Looks like this person is just using "ANA" as a synonym for what some of us here call Neo-African (of the pre-Eurasian bottleneck variety), which is to say the majority ancestry in Tropical Africans today (the non-Paleo-African, non-Eurasian portion). The term "ANA" should probably be reserved for the Neo-African ancestry in Taforalt (and related gene flow into Natufians). I wouldn't pay much attention to those estimates as some of them don't really make any sense (PPNB with twice as much ANA as Natufians? Yeah, no.)

El-Maestro
03-29-2021, 01:33 AM
Why is KEN_Kakapel at 87% and Dinka is at 74%? What's the main difference between these two populations?

Low SNP count in the former...

El-Maestro
03-29-2021, 01:36 AM
Looks like this person is just using "ANA" as a synonym for what some of us here call Neo-African (of the pre-Eurasian bottleneck variety), which is to say the majority ancestry in Tropical Africans today (the non-Paleo-African, non-Eurasian portion). The term "ANA" should probably be reserved for the Neo-African ancestry in Taforalt (and related gene flow into Natufians). I wouldn't pay much attention to those estimates as some of them don't really make any sense (PPNB with twice as much ANA as Natufians? Yeah, no.)

I'll ask a better question...

For the pattern below:
D(Mbuti x; Villabruna; test): Jordan_PPNB > Israel_PPNB
D(Mota x; Villabruna; test): Jordan_PPNB < Israel_PPNB
D(Taforalt x; Villabruna test): Jordan_PPNB > Israel_PPNB

Would it be "Neo Africans" responsible for this discrepancy? If not, then who?
I wan't you to keep in mind how the Natufians in the chart posted above relates to the Jordan PPNB samples and the the added fact that Natufians have been showing evidence of requiring Mota-related ancestry (https://anthrogenica.com/showthread.php?22253-African-phylogeny-(beginning-)&p=722283&viewfull=1#post722283) on top of "Basal Eurasian."

I can say that if there was a "Neo African" population in-line with what was being suggested here, Natufians should undoubtedly score more...

Brandon S. Pilcher
03-29-2021, 09:00 AM
Looks like this person is just using "ANA" as a synonym for what some of us here call Neo-African (of the pre-Eurasian bottleneck variety), which is to say the majority ancestry in Tropical Africans today (the non-Paleo-African, non-Eurasian portion). The term "ANA" should probably be reserved for the Neo-African ancestry in Taforalt (and related gene flow into Natufians). I wouldn't pay much attention to those estimates as some of them don't really make any sense (PPNB with twice as much ANA as Natufians? Yeah, no.)
What exactly does "Paleo-African" mean in the first place?

I can see PPNB having more ANA than Natufians if there were additional migrations out of Africa into the Levant after the Natufian period. Proto-Semitic would descend from one of these. Maybe what we're seeing with the Israeli PPNB reflects another wave of African migrations?

https://upload.wikimedia.org/wikipedia/commons/b/bc/BayesianSemiticMap.jpg

theplayer
03-29-2021, 09:34 AM
Typically "Paleo-African" would refer to a paraphyletic grouping of all modern humans except "Neo-African" and Eurasians. So "Khoisan" peoples would be >90% of that, around half of Central African Foragers' ancestry and somewhere between 10-40% of West Africans' and East Africans'. "Neo-African" would be the majority of ancestry in East and West Africans(and after the Bantu Migration also most of Subequatorial Africa).

Personally I would use Neo-African as the common ancestor of AEA, ANA and Eurasian and Paleo-African as the ancestor of all "non-archaic"(so likely excluding "Ghost Archaic" from the Shum Laka paper as well as Jebel Irhoud and Florisbad etc fossils) humans since I dislike using paraphyletic groupings(it makes it sound like they are more related than they actually are).

I don't think it is possible for PPNB to have more ANA than Natufians, even if the Afroasiatic-speakers from South-Eastern Egypt had more ANA than natufians, since PPNB would have more northern more Anatolian-like admixture far outweighing any excess ANA in Afroasiatic speakers. And that is if Semitic(-Berber) languages had even migrated to the Levant by that point but othesr might know more than me about the exact timing of that.


Also see the following F4-stats(Mor=Taforalt, Jor=PPNB, Isr=Natufian, Eth=Mota):
pop1 pop2 pop3 pop4 est se z p.value
1 Chi Mor Jor Isr 0.0026 0.0012 2.13 0.033443
2 Chi Eth Jor Isr 0.0008 0.0013 0.6 0.547355
3 Mbu Mor Jor Isr 0.0021 0.0008 2.45 0.014127
4 Mbu Eth Jor Isr 0.0002 0.001 0.23 0.815096

Natufian quite clearly has more ANA.

Edit: Just noticed that there are two different PPNB in the image(jordan/israel) of which only Israel_PPNB has more ANA, and I used Jordan_PPNB in the F4-stats....
Though I can't say I've ever heard of there being such a difference between different PPNB populations so it's kind of weird?
Edit2: Just checked in G25 since both PPNB and Natufian has contaminated outliers. That doesn't appear to be the problem (equal amounts of ANA in non/contaminated), so yeah weird.

El-Maestro
03-29-2021, 12:24 PM
Typically "Paleo-African" would refer to a paraphyletic grouping of all modern humans except "Neo-African" and Eurasians. So "Khoisan" peoples would be >90% of that, around half of Central African Foragers' ancestry and somewhere between 10-40% of West Africans' and East Africans'. "Neo-African" would be the majority of ancestry in East and West Africans(and after the Bantu Migration also most of Subequatorial Africa).

Personally I would use Neo-African as the common ancestor of AEA, ANA and Eurasian and Paleo-African as the ancestor of all "non-archaic"(so likely excluding "Ghost Archaic" from the Shum Laka paper as well as Jebel Irhoud and Florisbad etc fossils) humans since I dislike using paraphyletic groupings(it makes it sound like they are more related than they actually are).

I don't think it is possible for PPNB to have more ANA than Natufians, even if the Afroasiatic-speakers from South-Eastern Egypt had more ANA than natufians, since PPNB would have more northern more Anatolian-like admixture far outweighing any excess ANA in Afroasiatic speakers. And that is if Semitic(-Berber) languages had even migrated to the Levant by that point but othesr might know more than me about the exact timing of that.


Also see the following F4-stats(Mor=Taforalt, Jor=PPNB, Isr=Natufian, Eth=Mota):
pop1 pop2 pop3 pop4 est se z p.value
1 Chi Mor Jor Isr 0.0026 0.0012 2.13 0.033443
2 Chi Eth Jor Isr 0.0008 0.0013 0.6 0.547355
3 Mbu Mor Jor Isr 0.0021 0.0008 2.45 0.014127
4 Mbu Eth Jor Isr 0.0002 0.001 0.23 0.815096

Natufian quite clearly has more ANA.

Edit: Just noticed that there are two different PPNB in the image(jordan/israel) of which only Israel_PPNB has more ANA, and I used Jordan_PPNB in the F4-stats....
Though I can't say I've ever heard of there being such a difference between different PPNB populations so it's kind of weird?
Edit2: Just checked in G25 since both PPNB and Natufian has contaminated outliers. That doesn't appear to be the problem (equal amounts of ANA in non/contaminated), so yeah weird.

Do you think clear indications of additional "Paleo-" and "Neo-" African drift shared by Natufians highlight that the Argument which defines questioned admixture as "Neo-African" contradictory?
Look at what happens in the chart when more Ballitobay ancestry is deducted from Mota.

I think I can provide context for the difference in Jordan vs Israel.
Dstats:
Mota Chimp.REF Villabruna Jordan_PPNB 0.0026 0.445
Mota Chimp.REF Villabruna Israel_PPNB -0.0025 -0.456
Mbuti Chimp.REF Villabruna Jordan_PPNB 0.0026 0.541
Mbuti Chimp.REF Villabruna Israel_PPNB -0.0045 -1.059
Taforalt Chimp.REF Villabruna Jordan_PPNB -0.0125 -2.53
Taforalt Chimp.REF Villabruna Israel_PPNB -0.0243 -4.672
Taforalt Chimp.REF Israel_PPNB Jordan_PPNB 0.0138 2.685

Here's some more following your equations but also including Israel PPNB:
Chimp.REF Taforalt Israel_PPNB Natufian 0.006 1.145
Chimp.REF Mota Israel_PPNB Natufian -0.0088 -1.499
Mbuti Taforalt Israel_PPNB Natufian 0.0091 1.934
Mbuti Mota Israel_PPNB Natufian -0.0032 -0.666
Chimp.REF Taforalt Jordan_PPNB Natufian 0.0107 2.057
Chimp.REF Mota Jordan_PPNB Natufian -0.0011 -0.176
Mbuti Taforalt Jordan_PPNB Natufian 0.0121 2.822
Mbuti Mota Jordan_PPNB Natufian 0.0014 0.256

Also it isn't healthy to assume supposed "Known" Admixture from an unrelated source will dampen ancestry from another. For example Modern Nubians have Additional Non-African Admixture since the Christian Era. What'd that do to their putative African ancestry?
(https://t.co/eRCpVUJu3i?amp=1)