PCA G25_scaled - Africans only (Users & Ancients/Modern Avgs)

[Mnemonics]

Which East African population?

Dinka or Sudanese, Sudanese7 seems to be the most AEA like.

[Chad Rohlfsen]

Okay. Here is something new, that is in the works. It is kind of perplexing, but simple and really works. There aren't any issues with 0 (zero) drift-edges at all.

I am going to add in Natufians, Jomon, and ANE soon. After that will be Native Americans and the rest of the Near East, NorthAfrica, and pastoralists.

FullNew1.pngNewFullTop.pngNewFullBottom.png

[Chad Rohlfsen]

A bit of an update to it.

NewAfricaEurasia1Full.pngAfricaEurasiaNew1Top.pngAfricaEurasiaFull1Bottom.png

[K33]

Okay. Here is something new, that is in the works. It is kind of perplexing, but simple and really works. There aren't any issues with 0 (zero) drift-edges at all.

I am going to add in Natufians, Jomon, and ANE soon. After that will be Native Americans and the rest of the Near East, NorthAfrica, and pastoralists.

FullNew1.pngNewFullTop.pngNewFullBottom.png

Ehh, those are some pretty enormous drift paths required to connect the "OOA" node to the actual Paleolithic Eurasian samples underneath....




How does your fit/score compare to the one Kale put together using the find_graphs function? Kale's graph has 18% Basal Human admixture into Neanderthals... and that implies its from some early OOA hominid like Apidima-1.

I think we need to lose the idea that there was some orderly, single-file bottlenecked OOA exit out of East Africa circa 60-100kya. It seems much more complicated than that...

EDIT: just gonna post a jpeg of Kale's graph below... he reportedly got a score of 63.78 on this

[theplayer]

Couldn't the drift-edges' f2-stats differ by a scaling constant so you can't compare them directly?
I mean the f2 distances in qpgraph is scaled already(F2≤ 1 normally, in qpgraphs they are always bigger).

Or perhaps(likely) I am completey wrong on this.

[ThaYamamoto]

No problem at all! Also, I see what you mean now with what you were referring to above. I quite agree.



You certainly have a point, and thanks for dropping the link. In Extensive Admixture and Selective Pressure Across the Sahel Belt, that variation you're talking about is evident. In the study I previously mentioned, the West Sahelian/West Atlantic component is considerably greater amongst the Yoruba, for example, than in the study I just linked now. Rather odd, but I suppose one can only expect a considerable deal of variation depending on the study one employs. For instance, I've seen some studies show Ethiopian Jews as something like ~65%-70% West Eurasian on average (i.e., likely on par with some of the earliest Cushitic speakers’ Eurasian ancestry (Edit: to be specific, Natufian-like ancestry)), while some others list their West Eurasian ancestry as ~15%-30% lower, like with other Ethiosemitic speakers. When looking at Ethiopian Jews themselves, some of whom look rather like most unadmixed Sub-Saharan Africans (e.g., I've seen some with very dark skin and clearly Afro-textured hair rather than curly hair), I knew the ~70% figure was likely too high of an estimate, and I would scratch my head at how the researchers arrived at such a high figure, haha. Although, naturally, that’s just what their methods arrived at. I suppose this kind of variation even with regard to Niger-Congo-speaking populations lets us see each method's relative efficacy.

Yes the variation in results can be quite staggering particularly when first getting into this sort of thing, after a while you get used to it and learn what to look out for, supplemental information always help, haha. Would have to ask the East-African/Afro-Asiatic experts on here regarding this however, in terms of the total 'Eurasian' ancestry in HoAs, while their still needs to be a resolution as to the origins of both native and west-Eurasian ancestral components they harbor (and the phylogenies in general of Africa/Eurasia] I think 60% west-Eurasian for Ethiopian Jews/Amhara will be around the final figure when all is said and done. 70% is way too high indeed.

Thanks for the clarification. That makes a lot of sense what you're saying.

No problem!

Can you elaborate more on how the uniparentals were pointing you toward the Sahel/Mali region? I have noticed some signs of this myself specifically concerning uniparentals, but it would be interesting seeing how this has been the case for others.

Sure, just looking over the maternal lineages and their downstream frequencies in Bantu speaking regions and then cross-checking them against clades further upstream. For example L3b1a1/L3b1a1a dominates Bantu groups but a more westerly/Atlantic regions that are associated with high L3b frequencies (so far Senegal, Gambia, Guinea, Ivory Coast) are usually of a different branch altogether [L3b2 or specific downstream clades like L3b1b/L3b1a17/L3b1a-a*]. However L3b1a is found at high frequencies in Burkina Faso and among Fula groups. The same sort of pattern occurs with L3e, with non-Bantu clades often forming their own branches altough again, Sahelian groups like Burkinabe harbor the ancestral/upstream variant of the Bantu variant. Interestingly in terms of diversity the Burkinabe exhibit some of the highest mtdna haplotype diversity with both upstream Bantu and non-Bantu associated haplogroups of the same clade [L3b/L3e].So far these have been my loose observations and are far from conclusive, and my knowledge on L3e needs some revision, but this is just what I've come to find.

For the overestimation of Nigerian ancestry in your link, I get what you mean, but I don’t know much of an overestimation this is. I mean, in my case, I’ve only ever done AncestryDNA as my primary DNA test, and I get the impression that its “additional communities/migrations” section would best speak to what you’ve mentioned. Despite the fact that my background is Nigerian, I’ve only ever been grouped in with [Northern] Afro-Jamaican and Afro-Haitians, for instance. My DNA results were as follows (the first is the original):

• 39% Nigeria, 36% Benin/Togo, Cameroon/Congo, 19% Cameroon/Congo, 3% Senegal, <1% Ivory Coast/Ghana, <1% Africa South/Central Hunter-Gatherers, <1% Mali.

That was quite varied more varied than I thought, but I still found it interesting (especially the Senegal score, and “Nigeria” was still my highest percentage. Otherwise, the Benin/Togo percentage and the Cameroon/Congo scores were in line off a typical Nigerian, especially Southern ones.

The results of my first update were massively different to the point that I had trouble taking it seriously. It said this:

• 51% Benin/Togo, 44% Cameroon, Congo, and Southern Bantu peoples, 5% Nigeria (LOL!).

My next update said this, which was still massively different from the original, but similarly believable:

• 95% Nigeria, 5% Congo, and Southern Bantu peoples

My latest update was far simpler:

• 100% Nigeria

One can reasonably take the 100% score with a grain a salt, since it really is more like a simplification, but I do think it’s correct with what I know at the moment, which is that I haven’t been able to identify any roots from outside of Nigeria.

To return to the “additional communities/migrations” section provided by AncestryDNA, I think they’re somehow linking me to Afro-Jamaicans and Afro-Haitians, because the company somehow believes I resemble those populations the most (and vice versa) (i.e., maybe a supposed similar blend of ancestries. Also, for many African-Americans I’ve seen with the most elevated “Nigerian” percentages (50%+), they tend to have family roots in places where much of the enslaved populace was from what would become Nigeria (e.g., Virginia).

Plus, from the list of relatives on AncestryDNA, so far, contrary to what the “additional communities/migrations” section suggests, almost all my relatives have been exclusively African-American. One of the relatives I’ve gotten to know quite well actually has a large amount of identified Nigerian relatives besides me, and they’ve almost all been Igbo (the majority) or Yoruba (the slight minority). Otherwise, for a few of my African-American friends who’ve been tested, they actually don’t have “Nigeria” as their largest score. Nigeria will always seem to be the second largest region for them instead, but often, “Cameroon/Congo” or “Ivory Coast/Ghana” will be the prevailing region.

So, I get where you’re coming from, and who knows, maybe I’m ultimately misinterpreting the information, but I don’t get the impression that 23andme is necessarily artificially raising “Nigeria” scores. The “Nigeria” percentages I’ve seen for 23andme also seem to have been most often in line with AncestryDNA’s “Nigerian” scores for both African-Americans and Nigerians (at least the ones I’ve seen). If that weren’t the case, I think AncestryDNA would likely list me as something else, which would be strange, since I am Nigerian. The same could apply to my aforementioned relative, because he’s done both AncestryDNA and 23andme, and he’s gotten quite comparable “Nigeria” scores. That’s with the obvious caveat that “Nigerian” is a nationality first, not a genetic identity. Although, truthfully, my background in genetics is less robust than my understanding of linguistics. So, I could be missing something.

Thanks for sharing your scores, real interesting stuff - I've keenly been following the breakdowns with each successive update also. Looking back on my original post, I think I'd have to revise some, if not a lot, of what I first thought. My main contention was the disparity between IBD and Admixture proportions within that particular paper, but now I've been through the data again it seems I missed the fact they split Bight of Benin and Bight of Biafra into IBD streams but one ancestral admixture component, so this was an oversight on my part - apologies. I do now think that the marginal majority in most Aframs/Afro-Carribbeans is indeed modern day 'Nigerian' (as opposed to Benin as the Fon were notorious in their slave raiding and cultivation) followed by a Angolan/Congo component. As Cameroonians are intermediate I can see how this is still causing problems, ideally the paper should have included Cameroonian populations and perform an IBD analysis there. But yeah, I'm gonna put my hands up and say 'Nigerian' is the marginally dominant ancestry in AAs/ACs for sure, along with the coastal [Ivorian/Ghanaian] and Congolese/Cameroonian ancestries. I think I went a little hard with my prior assertions based off of what I'd seen with the IBD analyses, should have been a lot more tentative. Ancestry do seem to offer the more robust analyses at the moment.

In regards to AncestryDNA matches I have over 100 or so Afram as well as some Afro-Carribbeans and yes, one or two Afro-Haitians in my relative list. Without going off topic my great-grandmother herself claimed west-African ancestry and this has been the only indicator of there being some truth to this as, as far as I know, full-blooded SE Africans only tend to have a handful if any new world descendant matches, if at all. I plan to visit her clan village post covid and explore this further. The magic of IBD eh. I have a thread dedicated to this but I haven't updated it in a while, maybe I'll make a video showing a quick run through my matches. I did have 2% Nigerian on Ancestry prior to the update before this was shoe-horned into Cameroonian.

[gihanga.rwanda]

Yes the variation in results can be quite staggering particularly when first getting into this sort of thing, after a while you get used to it and learn what to look out for, supplemental information always help, haha. Would have to ask the East-African/Afro-Asiatic experts on here regarding this however, in terms of the total 'Eurasian' ancestry in HoAs, while their still needs to be a resolution as to the origins of both native and west-Eurasian ancestral components they harbor (and the phylogenies in general of Africa/Eurasia] I think 60% west-Eurasian for Ethiopian Jews/Amhara will be around the final figure when all is said and done. 70% is way too high indeed.

Based on their F4ratios, Eritreans (particularly Tigrinya speakers) are slightly more African than Taforalt and should be at least ~50% African or more precisely 50% Dinka/Mota vs. 50% proto-Natufian. Amhara and Ethiopian Jews are marginally less Eurasian falling between ~50-55%.

We need to wait for more ancient DNA from the region but I don’t think these admixture proportions (Africa vs. Eurasia) in Ethiopians etc. will drastically change. But who knows, Lipson et al. 2020 modeled Agaws (i.e. Ethiopian Jews) as ~60% African but that particular model didn’t account for Basal Eurasian ancestry.

[Angoliga]

"ZAF_Zulu_Sotho_Swati_N", "ZAF_Zulu_Swati_T" samples added:


Updated All-African PCA: (Total = 212 samples)

 

(North-African, Tropical-African, South-African)

◇ (50) = Ancients
X (21) = Users
• (141) = Moderns

3D PCA: https://plot.ly/~Angoliga/3/ (allow ~less than a minute to load)
2D PCA: https://plot.ly/~Angoliga/4/

Deep-layer results:

Target: ZAF_Zulu_Sotho_Swati_N
Distance: 4.9783% / 0.04978327
68.6 Yoruba
19.6 ZAF_2000BP
11.2 Sudanese
0.6 Levant_Natufian

Target: ZAF_Zulu_Swati_T
Distance: 4.7617% / 0.04761723
67.8 Yoruba
18.4 ZAF_2000BP
12.2 Sudanese
1.6 Anatolia_Barcin_N

Source samples:

 

Levant_Natufian:I0861,0.01935,0.135065,-0.039221,-0.135984,0.026774,-0.076137,-0.019036,-0.024691,0.100626,-0.008018,0.02858,-0.019633,0.067343,0.001651,0.022801,0.02612,-0.0103,0.006714,-0.018101,0.041395,-0.004118,-0.003215,-0.014297,-0.011206,0.011975
CHN_Tianyuan,-0.027318,-0.260991,-0.075424,0.071383,0.033545,-0.018407,-0.00799,-0.003,0.040291,0.021322,-0.006333,-0.005995,-0.003568,-0.00523,-0.000407,-0.000663,0.008084,0.002407,-0.001131,0.027263,-0.001747,0.008037,-0.008874,-0.010845,0.011256
IND_Great_Andamanese_100BP:Andaman,-0.018212,-0.235603,-0.128598,0.098838,0.023389,-0.003904,-0.016686,-0.003231,0.059721,0.02041,0.023222,0.004646,-0.0055,0.001239,-0.016151,-0.008884,0.009518,0.001014,-0.004902,0.026888,-0.00549,0.011994,-0.013311,-0.001205,0.000958
Yamnaya_UKR,0.119514,0.088351,0.044123,0.110144,-0.032006,0.038487,0.00799,0.002308,-0.05788,-0.076721,-0.002761,-0.002398,0.000743,-0.021056,0.033387,0.01485,0.004955,-0.008235,-0.002263,0.015007,0.002121,0.003462,0.002958,0.023 256,0.002874
Anatolia_Barcin_N,0.1175998,0.180118,0.0035312,-0.101158,0.0510443,-0.0483875,-0.0043582,-0.0069334,0.0362287,0.0807473,0.0079718,0.0118803,-0.0234545,0.0004691,-0.0419807,-0.0101913,0.0233091,0.0019866,0.0136954,-0.0097489,-0.0142249,0.0057723,-0.0041232,-0.0031658,-0.0043437
Levant_Natufian,0.020488,0.1431895,-0.0377125,-0.1387295,0.030775,-0.079484,-0.025616,-0.0175375,0.114329,0.002005,0.0332085,-0.0222555,0.076486,0.002133,0.0153365,0.009016,-0.0154505,-0.001014,-0.02206,0.040832,0.001497,0.0001235,-0.003636,-0.0044585,0.006287
Yoruba,-0.6300625,0.0625011,0.022113,0.0167079,0.0005035,0 .0124741,-0.044417,0.0477673,-0.0488813,0.0327694,0.0046205,0.0007904,0.0230561, 0.0009509,0.0125232,-0.0096067,0.0070763,0.0004491,0.006022,-0.00299,0.0015542,0.0023156,-0.0017592,-0.0004711,-0.0004246
ZAF_2000BP,-0.641583,0.0558543,0.0258957,0.034884,0.001231,-0.0030677,0.2680687,-0.2072223,0.0104307,0.0153683,0.008715,-0.0659913,-0.0216053,0.0064223,0.023751,-0.0201537,0.034204,0.294593,-0.09574,0.0060447,-0.033857,-0.0039157,0.003163,-0.0029323,-0.000399
Sudanese:SUDANESE7,-0.571392,0.046714,-0.007165,-0.007429,-0.005232,-0.002789,-0.00799,0.018922,0.0904,-0.108977,-0.018512,0.020831,-0.047125,0.000688,0.013165,-0.023734,0.021644,-0.012035,0.026145,-0.023261,-0.000998,0.004822,-0.000986,-0.001687,0.003712
Mbuti,-0.6493608,0.0578852,0.0250785,0.038437,-0.001077,0.0054385,0.3631505,-0.291911,0.0193278,0.0323013,-0.0090938,0.2169692,0.1242805,3.45E-05,0.0080072,0.0017568,0.000717,-0.049567,0.015901,0.004471,0.0049288,-0.0020402,-0.002465,0.0057838,0.000389
IRN_Wezmeh_N,0.037562,0.072103,-0.165556,-0.016473,-0.11756,0.01506,0.017861,-0.003231,-0.071583,-0.046835,0.003085,-0.003147,0.00669,0.000688,0.022122,0.053036,-0.011735,0.001014,0.014581,-0.035267,0.009358,-0.023741,-0.006655,-0.032896,0.019998
HUN_Koros_N_HG,0.137726,0.12491,0.180264,0.154395, 0.128332,0.042112,0.015041,0.03646,0.071992,0.0069 25,-0.005196,-0.013488,0.014717,-0.003303,0.028637,0.046141,0.018384,0.005954,-0.008296,0.045772,0.061142,0.010634,-0.040672,-0.124475,0.013053
GEO_CHG,0.091058,0.102568,-0.083344,-0.00323,-0.08617,0.020638,0.024911,-0.001846,-0.128236,-0.074717,-0.006333,0.023979,-0.054856,0.004404,0.026601,-0.03275,0.02386,-0.013429,-0.022249,0.034767,0.033815,-0.007048,0.006532,-0.025787,-0.002036
MAR_Iberomaurusian,-0.189857,0.0814452,-0.0242866,-0.085595,0.027636,-0.0552202,-0.0705968,0.0184146,0.155397,0.003499,0.0209156,-0.0318316,0.0747168,-0.0513334,0.0711988,-0.0363032,0.0052676,-0.066106,-0.1424162,0.0389938,-0.0376836,-0.1255322,0.0730118,-0.0137606,0.0164534

[drobbah]

How did these South Africans get Sudanese ancestry?

[theplayer]

Target	Distance	ZAF_2000BP	Sudanese	Yoruba	...	Distance	ZAF_2000BP	Sudanese	Kongo	...	Distance	ZAF_2000BP	Sudanese	Manyika
Bantu_SE	4.29%	21.8	14.2	64.0	...	2.22%	17.4	4.8	77.8	...	1.05%	16.2	2.6	81.2