Niger-Congo Ancestry In East African Peoples
#16
I'm not sure what kind of Vahaduo you are using, but in original Vahaduo your distances should be multiplied by 100.

By the way, this is what East African peoples which you selected score in my 10 Thousand Years Ago calculator:

Target: Luo_(Kenya_&_Tanzania)
Distance: 6.2393% / 0.06239314
41.6 West_Africa_10KYA
38.6 Saharan_Africa_10KYA
13.4 Central_Africa(west)_10KYA
4.6 Northwest_Africa_10KYA
1.8 Anatolia_Greece_10KYA

Target: Agaw_(Ethiopia)
Distance: 4.6069% / 0.04606899
55.2 Levant_10KYA
35.4 Saharan_Africa_10KYA
7.8 Northeast_Africa_10KYA
1.6 Caucasus_10KYA

Target: Afar_(Ethiopia)
Distance: 4.6318% / 0.04631792
54.4 Levant_10KYA
40.4 Saharan_Africa_10KYA
2.4 Zagros_Mountains_10KYA
1.8 Northeast_Africa_10KYA
0.8 Steppe_Proto_Yamnaya
0.2 Caucasus_10KYA

Target: Somali_(Ethiopia)
Distance: 4.9843% / 0.04984300
51.0 Saharan_Africa_10KYA
45.0 Levant_10KYA
3.2 Northeast_Africa_10KYA
0.8 Northwest_Africa_10KYA

Target: Ari.Cultivator_(Ethiopia)
Distance: 2.8469% / 0.02846915
54.8 Northeast_Africa_10KYA
23.2 Saharan_Africa_10KYA
21.4 Levant_10KYA
0.6 Northwest_Africa_10KYA

Target: Ari.blacksmith_(Ethiopia)
Distance: 2.6828% / 0.02682758
56.2 Northeast_Africa_10KYA
23.6 Saharan_Africa_10KYA
19.4 Levant_10KYA
0.8 Northwest_Africa_10KYA

Target: Ari_(Ethiopia)
Distance: 3.2929% / 0.03292862
59.4 Northeast_Africa_10KYA
20.0 Levant_10KYA
19.8 Saharan_Africa_10KYA
0.8 Northwest_Africa_10KYA

Target: Anuak_(Ethiopia)
Distance: 2.1969% / 0.02196946
82.2 Saharan_Africa_10KYA
12.4 Northeast_Africa_10KYA
3.8 West_Africa_10KYA
1.4 Central_Africa(west)_10KYA
0.2 Levant_10KYA

Target: Wolayta_(Ethiopia)
Distance: 3.7425% / 0.03742468
40.2 Levant_10KYA
33.6 Saharan_Africa_10KYA
25.0 Northeast_Africa_10KYA
0.8 Zagros_Mountains_10KYA
0.4 Northwest_Africa_10KYA

Target: Mursi_(Ethiopia)
Distance: 2.8501% / 0.02850143
71.0 Saharan_Africa_10KYA
23.0 Northeast_Africa_10KYA
6.0 Levant_10KYA

Target: Sandawe_(Tanzania)
Distance: 2.0651% / 0.02065094
38.0 Saharan_Africa_10KYA
21.8 East_Africa_10KYA
17.4 Levant_10KYA
9.6 West_Africa_10KYA
5.8 South_Africa_10KYA
5.0 Central_Africa(west)_10KYA
1.8 Northwest_Africa_10KYA
0.6 Zagros_Mountains_10KYA

Target: Hadza_(Tanzania)
Distance: 0.5539% / 0.00553912
31.8 Northeast_Africa_10KYA
21.0 Saharan_Africa_10KYA
20.6 East_Africa_10KYA
12.4 Central_Africa(west)_10KYA
6.6 Levant_10KYA
6.0 Central_Africa(east)_10KYA
0.6 Southern_Asia_10KYA
0.4 Scandinavia_10KYA
0.4 Western_Iran_10KYA
0.2 Steppe_Proto_Yamnaya

Target: Gumuz_(Ethiopia)
Distance: 2.0933% / 0.02093285
62.6 Saharan_Africa_10KYA
32.6 Northeast_Africa_10KYA
4.6 Levant_10KYA
0.2 Western_Iran_10KYA

Target: Dinka_(South.Sudan)
Distance: 1.4217% / 0.01421713
92.4 Saharan_Africa_10KYA
4.6 Northeast_Africa_10KYA
3.0 West_Africa_10KYA

In my calculator Mota-like ancestry peaks in Omotic Ari people, not in Hadza.

And Dinka are the purest (most Proto-Nilotic) of existing Nilotic populations.
Nguni likes this post
Reply
#17
(12-31-2024, 02:43 AM)Tomenable Wrote: So far there are no ancient Nilotic samples available.

Ancient Mota was not a Nilotic sample, it was Omotic. This is why modeling the Dinka with Mota produces such a bad distance.

I know that Mota was not Nilotic. However, he also was not Omotic; rather, people like him were the source of about half of the ancestry in Omotic peoples, as Omotic peoples are part Mota-like and part Cushitic (Dinka-like and Natufian), Arabian (in the case of the Wolayta), and part North African (as represented by Ancient Egyptian and Morocco Neolithic samples).

   
  • Omotic peoples = Mota-like + Cushitic
  • Cushitic = Dinka-like + Natufian

In my opinion, the following occurred:
  • Natufians migrated into Egypt ~12 kya.
  • Some later migrated into Sudan and mixed with Proto-Nilotes, thus creating Proto-Cushites.
  • Proto-Cushites migrated into the Horn of Africa, wherein some of them mixed with Mota-like hunter-gatherers and created Omotic peoples.
  • Ancient Egyptians migrated into the Horn of Africa ~3 kya and were absorbed by Cushites and some Omotic peoples.
  • Arabians migrated into the Horn of Africa sometime after the birth of Islam and were absorbed by Cushitic and Omotic peoples. 

By the way, below is a visual representation of my hypotheses regarding the origin of modern lineages.

[Image: attachment.php?aid=2559]
Necrontyr, Mythbuster General, Tomenable like this post
Reply
#18
(12-31-2024, 03:16 AM)Tomenable Wrote: I'm not sure what kind of Vahaduo you are using, but in original Vahaduo your distances should be multiplied by 100.

By the way, this is what East African peoples which you selected score in my 10 Thousand Years Ago calculator:

Target: Luo_(Kenya_&_Tanzania)
Distance: 6.2393% / 0.06239314
41.6 West_Africa_10KYA
38.6 Saharan_Africa_10KYA
13.4 Central_Africa(west)_10KYA
4.6 Northwest_Africa_10KYA
1.8 Anatolia_Greece_10KYA

Target: Agaw_(Ethiopia)
Distance: 4.6069% / 0.04606899
55.2 Levant_10KYA
35.4 Saharan_Africa_10KYA
7.8 Northeast_Africa_10KYA
1.6 Caucasus_10KYA

Target: Afar_(Ethiopia)
Distance: 4.6318% / 0.04631792
54.4 Levant_10KYA
40.4 Saharan_Africa_10KYA
2.4 Zagros_Mountains_10KYA
1.8 Northeast_Africa_10KYA
0.8 Steppe_Proto_Yamnaya
0.2 Caucasus_10KYA

Target: Somali_(Ethiopia)
Distance: 4.9843% / 0.04984300
51.0 Saharan_Africa_10KYA
45.0 Levant_10KYA
3.2 Northeast_Africa_10KYA
0.8 Northwest_Africa_10KYA

Target: Ari.Cultivator_(Ethiopia)
Distance: 2.8469% / 0.02846915
54.8 Northeast_Africa_10KYA
23.2 Saharan_Africa_10KYA
21.4 Levant_10KYA
0.6 Northwest_Africa_10KYA

Target: Ari.blacksmith_(Ethiopia)
Distance: 2.6828% / 0.02682758
56.2 Northeast_Africa_10KYA
23.6 Saharan_Africa_10KYA
19.4 Levant_10KYA
0.8 Northwest_Africa_10KYA

Target: Ari_(Ethiopia)
Distance: 3.2929% / 0.03292862
59.4 Northeast_Africa_10KYA
20.0 Levant_10KYA
19.8 Saharan_Africa_10KYA
0.8 Northwest_Africa_10KYA

Target: Anuak_(Ethiopia)
Distance: 2.1969% / 0.02196946
82.2 Saharan_Africa_10KYA
12.4 Northeast_Africa_10KYA
3.8 West_Africa_10KYA
1.4 Central_Africa(west)_10KYA
0.2 Levant_10KYA

Target: Wolayta_(Ethiopia)
Distance: 3.7425% / 0.03742468
40.2 Levant_10KYA
33.6 Saharan_Africa_10KYA
25.0 Northeast_Africa_10KYA
0.8 Zagros_Mountains_10KYA
0.4 Northwest_Africa_10KYA

Target: Mursi_(Ethiopia)
Distance: 2.8501% / 0.02850143
71.0 Saharan_Africa_10KYA
23.0 Northeast_Africa_10KYA
6.0 Levant_10KYA

Target: Sandawe_(Tanzania)
Distance: 2.0651% / 0.02065094
38.0 Saharan_Africa_10KYA
21.8 East_Africa_10KYA
17.4 Levant_10KYA
9.6 West_Africa_10KYA
5.8 South_Africa_10KYA
5.0 Central_Africa(west)_10KYA
1.8 Northwest_Africa_10KYA
0.6 Zagros_Mountains_10KYA

Target: Hadza_(Tanzania)
Distance: 0.5539% / 0.00553912
31.8 Northeast_Africa_10KYA
21.0 Saharan_Africa_10KYA
20.6 East_Africa_10KYA
12.4 Central_Africa(west)_10KYA
6.6 Levant_10KYA
6.0 Central_Africa(east)_10KYA
0.6 Southern_Asia_10KYA
0.4 Scandinavia_10KYA
0.4 Western_Iran_10KYA
0.2 Steppe_Proto_Yamnaya

Target: Gumuz_(Ethiopia)
Distance: 2.0933% / 0.02093285
62.6 Saharan_Africa_10KYA
32.6 Northeast_Africa_10KYA
4.6 Levant_10KYA
0.2 Western_Iran_10KYA

Target: Dinka_(South.Sudan)
Distance: 1.4217% / 0.01421713
92.4 Saharan_Africa_10KYA
4.6 Northeast_Africa_10KYA
3.0 West_Africa_10KYA

In my calculator Mota-like ancestry peaks in Omotic Ari people, not in Hadza.

And Dinka are the purest (most Proto-Nilotic) of existing Nilotic populations.

1. I used Vahaduo's "Multi" mode to calculate those distances; it generates models for multiple Targets at once and groups them in a table. Also, it expresses percentages as decimal values (e.g. 10% = 0.1, 5% = 0.05, etc). So, 0.07 to 0.1 = 7% to 10%.

2. Vahaduo's distance calculator indicates that the Hadza are the people who are genetically closest to Mota, which indicates that they have more Mota-like ancestry than the Ari people.

   
Reply
#19
Dinka =/= Mota + Yoruba in a strict sense.
Chimp.REF Han.DG Dinka.DG Mota.SG -0.00175981 0.000357557 -4.92175 8.57718e- 7 1101092
Chimp.REF Han.DG Dinka.DG Yoruba.DG -0.00431671 0.000214568 -20.1181 5.12374e-90 1101471

Chimp.REF Cameroon_ShumLaka_8000BP.DG Dinka.DG Mota.SG -0.00335478 0.000311348 -10.7750 4.51575e-27 1092178
Chimp.REF Cameroon_ShumLaka_8000BP.DG Dinka.DG Yoruba.DG -0.00182929 0.000189018 -9.67787 3.74428e-22 1092051

Chimp.REF Congo_Mbuti.DG Dinka.DG Mota.SG -0.00304379 0.000226414 -13.4435 3.36346e-41 1101210
Chimp.REF Congo_Mbuti.DG Dinka.DG Yoruba.DG -0.00184710 0.000132416 -13.9492 3.18053e-44 1101548

Chimp.REF SouthAfrica_2000BP.SG Dinka.DG Mota.SG 0.000209781 0.000237220 0.884332 3.76517e- 1 1101612
Chimp.REF SouthAfrica_2000BP.SG Dinka.DG Yoruba.DG -0.00110532 0.000138301 -7.99211 1.32645e-15 1101431

On the other hand, that's not to say there's no connection.
Yoruba.DG
Dinka.DG 0.581365 0.0274812 21.1550
Cameroon_ShumLaka_8000BP.DG 0.373653 0.0248948 15.0093
Taforalt 0.0327972 0.00539424 6.08005
Denisova.DG 0.0121848 0.00223933 5.44126 (to stand in for extra 'A00' ancestry)
Tail = 6e-4
right = c('Chimp.REF', 'SouthAfrica_2000BP.SG', 'Mbuti.DG', 'Mota.SG', 'Morocco_OUB002_Epipaleolithic.SG', 'Anatolia_Barcin_N.SG')
The tail is bad, the residuals aren't in outer space though (worst outlier: Z = 2.32), so it should at least be close, Denisova obviously is an imperfect proxy, so that may be a culprit.

Also this is sort of interesting.
Cameroon_ShumLaka_8000BP.DG
Mbuti.DG 0.606629 0.0230218 26.3502
Yoruba.DG 0.393371 0.0230218 17.0869
Tail: 0.13
right = c('Chimp.REF', 'SouthAfrica_2000BP.SG', 'Mota.SG', 'Morocco_OUB002_Epipaleolithic.SG', 'Anatolia_Barcin_N.SG')

Mbuti.DG
Cameroon_ShumLaka_8000BP.DG 0.676932 0.0178822 37.8551
SouthAfrica_2000BP.SG 0.323068 0.0178822 18.0665
Tail: 0.15
right = c('Chimp.REF', 'Malawi_Hora_8500BP.SG', 'Mota.SG', 'Morocco_OUB002_Epipaleolithic.SG', 'Anatolia_Barcin_N.SG')

The more distal versions fail (Mbuti/ShumLaka = Yoruba + SouthAfrica). I'm thinking the two have a Yoruba-like + SouthAfrica-like mix of ancestries, but that those two proxies are only close enough to work at 30-40%, not good enough to work at 100%. That and/or the proxies are fine, but they have some degree of Central African ancestry (which I'm not sure whom would be the best representative)

The best I can figure is this, first just to associate the autosomal and uniparentals.
Ghost modern = A00
South-African = A1b1
Central African = B
ANEA = D0, E

Dinka/Mota = ANEA + Central + South. Mota has slightly less ANEA.
Niger-Congo = Ghost + Central + Dinka (or Dinka-related ANEA)
ShumLaka/Mbuti = Niger-Congo + South and/or Central
Tomenable, kolompar, Nguni And 1 others like this post
Reply
#20
The most "Denisovan-like" is baa001 / South_Africa_1900BP.SG / .
Yoruba-like - there are many of them and they seems to be late decendants.

Yoruba-like representatives are the most distant from Denisova / Neanderthals. They accumulate the most variety of non-Denisovan variants.

Other good representatives are Khomani San, Mbuti. ( which have some mix between Denisova-like , similar to baa001 and Yoruba-like)

The Eastern Africans have some mix between South-African + Yoruba-like. And Eastern Africans have the connection to the OOA initial group which are the ancestors of all Eurasians and non-africans in general.
Reply
#21
(12-31-2024, 03:38 AM)Inquirer Wrote: 2. Vahaduo's distance calculator indicates that the Hadza are the people who are genetically closest to Mota, which indicates that they have more Mota-like ancestry than the Ari people.

2. Not necessarily, closer distance doesn't always indicate that you have more of such ancestry. Because it depends on what you are mixed with. For example someone who is 50% Mota and 50% East Asian will have a farther distance to Mota than someone who is 40% Mota and 60% Niger-Congo. Simply because East Asian is more foreign.

Ari are mixed with Levantine which is Non-African and this is why they have farther distance to Mota, despite having more of Mota ancestry than Hadza.

(12-31-2024, 03:38 AM)Inquirer Wrote: 1. I used Vahaduo's "Multi" mode to calculate those distances; it generates models for multiple Targets at once and groups them in a table. Also, it expresses percentages as decimal values (e.g. 10% = 0.1, 5% = 0.05, etc). So, 0.07 to 0.1 = 7% to 10%.

1. Okay but a distance of over 10% (over 0.10) in your model is really bad, it indicates that your model for the Dinka is not realistic.

I think that the Dinka are their own group, not a mixture of Mota and Yoruba. But they are intermediate between Mota and Yoruba.
Reply
#22
In a PCA, Dinka don't plot exactly between Mota and West African, but instead they occupy their own, drifted position:

[Image: GZlf1vz.png]

I used this Custom PCA feature:

https://vahaduo.github.io/g25views/#Custom
Reply
#23
(12-31-2024, 03:26 AM)Inquirer Wrote: By the way, below is a visual representation of my hypotheses regarding the origin of modern lineages.

[Image: attachment.php?aid=2559]

^^^
Very interesting!

Are you sure that the Pygmies are an older lineage than the Khoisan?

This below is from my old notes (from 2017, so might be outdated now):

The Khoisan (Capoids) started to diverge around 160-150,000 years ago.

The Pygmies (Bambutids) started to diverge around 130,000 years ago.

Ancestral West Africans started to diverge around 90-65,000 years ago.

I have no notes about the time of divergence of Ancestral East Africans. I also have no notes about the lineage leading to the Taforalt (NW Africans).

==========

Now when it comees to the Out-of-Africa populations, from my 2017 notes:

South Eurasians (AASI, Negritos) started to diverge 65-55,000 years ago.

Sahulians (Australians, Papuans) started to diverge 60-40,000 years ago.

East Asians started to diverge 45,000-30,000 years ago.

West Eurasians started to diverge 45,000-30,000 years ago.

Native Americans started to diverge 30,000-20,000 years ago in Beringia.

Would you suggest any changes to these dates?

==========

If you make a PCA using populations from my Basic World K10 calculator, you will see that South Eurasians, Sahulians and East Asians* form three separate clusters:

https://www.exploreyourdna.com/calculato...cinski.htm

https://vahaduo.github.io/g25views/#Custom

[Image: o2BcQTV.png]

So I'm not sure if grouping all of these 3 clusters under the label "East Eurasians" makes sense. Maybe in a very distant past, shortly after OoA, they were one cluster.

*I labelled East Asians as East Eurasians in this calculator because they include also Lapita populations from Oceania, as well as North Asian populations from Russia.
Reply
#24
SouthAfricans are overall ~equally related to Mbuti and Sardinian, and more related to either than to Yoruba.
Chimp.REF SouthAfrica_2000BP.SG Yoruba.DG Sardinian.DG 0.00105542 0.000191466 5.51231 3.54148e-8 1101376
Chimp.REF SouthAfrica_2000BP.SG Mbuti.DG Sardinian.DG 0.000225725 0.000204650 1.10298 2.70036e-1 1101567

So unless Mbuti have South-Africa specific ancestry to offset any pre-South-Africa divergent ancestry, there's no reason to suspect they have any major such.
Reply
#25
(12-31-2024, 05:41 PM)Kale Wrote: SouthAfricans are overall ~equally related to Mbuti and Sardinian, and more related to either than to Yoruba.
Chimp.REF SouthAfrica_2000BP.SG Yoruba.DG Sardinian.DG 0.00105542  0.000191466 5.51231 3.54148e-8 1101376
Chimp.REF SouthAfrica_2000BP.SG Mbuti.DG    Sardinian.DG 0.000225725 0.000204650 1.10298 2.70036e-1 1101567

So unless Mbuti have South-Africa specific ancestry to offset any pre-South-Africa divergent ancestry, there's no reason to suspect they have any major such.

Well in PCA and G25 distances, Mbuti are closer to other Africans than to Sardinians, which makes sense.
Reply
#26
(12-31-2024, 05:29 PM)Tomenable Wrote:
Show Content

According to Figure S3.17 of Lipson et al (2020), a large minority of the ancestry in Niger-Congo peoples is Ghost Modern, and the shared source of ancestry in Pygmies is Ghost Modern related.

The Ghost Modern branch of modern humans preceded the Ancient Southern African branch, according to the figure.

   

I refer to the branch that leads to Pygmies as "Ghost Modern related" because it has 10 units of drift while that of Niger-Congo peoples has none. Additionally, other figures from the article, such as Figure 5, only label ancestry in Niger-Congo peoples as "Ghost modern."

   

The Ghost Modern branch that leads to Niger-Congo peoples indicates the following:

Niger-Congo peoples: 28.52% (92% of 31%) Ghost Modern 

The Ghost Modern related branches that lead to Shum Laka, Aka, and Mbuti indicate the following:

Shum Laka: 36.63% (37% of 99%) Ghost Modern related
Aka: 42% Ghost Modern related
Mbuti: 59.22% (63% of 94%) Ghost Modern related

Note that some Niger-Congo peoples and a Shum Laka forager carry / carried yDNA A00, which likely derived from their Ghost Modern (related) ancestry and which is far older than yDNA A1b1 of Khoisan peoples.
Tomenable likes this post
Reply
#27
(12-31-2024, 05:29 PM)Tomenable Wrote:
Show Content


Concerning Ancestral East Africans, I think that they would have been the remaining Anatomically Modern Humans in (Central) East Africa after the ancestors of Ancient Southern Africans / Khoisan peoples split from them. I believe that yDNA A1b1 and mtDNA L0 would have formed in Ancient Southern Africans; therefore, the relatives of these haplogroups, yDNA BT and mtDNA L3'4'6'2'5'1, would have formed in the remaining Anatomically Modern Humans of (Central) East Africa. Older subclades of yDNA A0-T (namely A0 & A1a) and unknown subclades of mTDNA L would have formed in (Central) East Africa, in my opinion, and spread in other directions.

Therefore, the remaining Anatomically Modern Humans of (Central) East Africa would have carried yDNA BT & mtDNA L3'4'5'2'5'1, and they are whom I refer to as Ancetral East Africans; their yDNA and mtDNA haplogroups have been estimated to have formed ~130 kya and ~160 kya, respectively. I use the younger date of ~130 kya as the latest time by which they would have formed as a distinct lineage; lineages that would have formed between ~160 kya and ~130 kya would have been carried out of the region, in my opinion, possibly by the ancestors of the Skhul and Qafzeh hominins of Israel and the Aterian hominins of North Africa.

Ancestral Northeast Africans (yDNA CT & L3) would have split from Ancestral East Africans ~88 kya, which is the estimative date for the formations of CT & the split of pre-L3 from L3'4.

The remaining Ancestral East Africans would have carried yDNA B-M60 and mtDNA L5, L6, & L4; mtDNA L1 & L2 would have formed in West / West-Central Africa due to lineages of Ancestral East Africans migrating thereto and mixing with Ghost Modern peoples and the Ghost Modern related ancestors of Pygmies. Millennia later, a subset of Ancestral Northeast Africans would have encountered Ghost Modern peoples in the Sahel or nearby, thereby creating Niger-Congo peoples in the Western Sahel (which would be why the latter carry subclades of yDNA D0, E-M96 and mtDNA L1, L2, and L3) and the shared ancestors of Nilotic and Mota-like peoples in the Eastern Sahel.

The shared ancestors of Nilotic and Mota-like peoples would have admixed with the remaining Ancestral East Africans. This would explain why Mota has Ghost Modern ancestry and why he's closely related to the Hadza people despite having very different haplogroups; this would also explain why Nilotic peoples have an affinity with Niger-Congo peoples (assuming that there wasn't an admixture event between them during the latest Green Sahara period).
Necrontyr and Tomenable like this post
Reply
#28
(12-31-2024, 05:41 PM)Kale Wrote: SouthAfricans are overall ~equally related to Mbuti and Sardinian, and more related to either than to Yoruba.
Chimp.REF SouthAfrica_2000BP.SG Yoruba.DG Sardinian.DG 0.00105542  0.000191466 5.51231 3.54148e-8 1101376
Chimp.REF SouthAfrica_2000BP.SG Mbuti.DG    Sardinian.DG 0.000225725 0.000204650 1.10298 2.70036e-1 1101567

So unless Mbuti have South-Africa specific ancestry to offset any pre-South-Africa divergent ancestry, there's no reason to suspect they have any major such.

Stats like the above sometimes make call into question the output of f4stats,  can't wrap my head around San sharing more drift with non-African populations like Sardinians than with other Sub-Saharan Africans such as Yoruba or Mbuti, especially considering things like PCA, Fst, or STRUCTURE runs at K = 2 always seem to indicated otherwise. It's not even limited to just San, I believe a stat such as (Chimp, Dinka, non-African, Yoruba) has a Z score of something like -20, which seems crazy to me. I get that Africans can be deeply structured but, some of these results would seem to indicate that theres been little mixing between any of them both before and after the OOA event
Reply
#29
(01-01-2025, 10:25 PM)ModusOperandi Wrote:
(12-31-2024, 05:41 PM)Kale Wrote: SouthAfricans are overall ~equally related to Mbuti and Sardinian, and more related to either than to Yoruba.
Chimp.REF SouthAfrica_2000BP.SG Yoruba.DG Sardinian.DG 0.00105542  0.000191466 5.51231 3.54148e-8 1101376
Chimp.REF SouthAfrica_2000BP.SG Mbuti.DG    Sardinian.DG 0.000225725 0.000204650 1.10298 2.70036e-1 1101567

So unless Mbuti have South-Africa specific ancestry to offset any pre-South-Africa divergent ancestry, there's no reason to suspect they have any major such.

Stats like the above sometimes make call into question the output of f4stats,  can't wrap my head around San sharing more drift with non-African populations like Sardinians than with other Sub-Saharan Africans such as Yoruba or Mbuti, especially considering things like PCA, Fst, or STRUCTURE runs at K = 2 always seem to indicated otherwise. It's not even limited to just San, I believe a stat such as (Chimp, Dinka, non-African, Yoruba) has a Z score of something like -20, which seems crazy to me. I get that Africans can be deeply structured but, some of these results would seem to indicate that theres been little mixing between any of them both before and after the OOA event

Vahaduo gives the impression that all Aboriginal Africans are more closely related to one another than they are to Non-Africans. Why do you think that Vahaduo gets this right (in my opinion) while f4stats do not? 

   
Reply
#30
- The vast majority of samples in G25 are Eurasian.
- The Eurasian bottleneck was extensive.
- PCA formulates dimensions which get the most 'bang for the buck', which roughly means # of samples multiplied by amount of drift.
Consequently Eurasian diversity is much more represented on G25.

A similar principle applies to admixture. IIRC at K2 you see an 'African' component and a 'Eurasian' component. The African populations closer to the OOA bottleneck get increasing levels of the 'Eurasian' component.

The direct f4-stats aren't showing a lack of mixing between Africans, quite the contrary, it's showing that there have been multiple mixtures between multiple deeply distinct layers of ancestry.
Look at West African y-hg for example, A00, A0, A1a, A1b1, B, D0, and of course E.

South-Africa seems more like an exception, correlating with the more constrained uniparentals (mt-L0, y-hg A1b1)
Chimp.REF SouthAfrica_2000BP.SG Sardinian.DG JuHoanNorth.DG 0.0106440 0.000234200 45.4486 0 1100820
ModusOperandi and theplayer like this post
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)