Ancient genomes from southern Africa

[LeBrok]

A less racist interpretation would be that common archaeological reference points are greatly underestimated and mutation rates overestimated. It's what happens when the most important places in the prehistory of humanity are understudied compared to the West Eurasian sink.

I thought that my joke was obvious, did you miss the blink at the end? You should know me better by now.

 

[berun]

I can't accept admixture analysis with recently mixed population, that is a basic rule (and much less to accept dating our spice with them!).

The Oorlams Edit
In the 19th century white farmers, mostly Boers, moved farther north, pushing the indigenous Khoisan peoples, who put up a fierce resistance, across the Orange River. Known as Oorlams, these Khoisan adopted Boer customs and spoke a language similar to Afrikaans.[2] Armed with guns, the Oorlams caused instability as more and more came to settle in Namaqualand and eventually conflict arose between them and the Nama. Under the leadership of Jonker Afrikaner, the Oorlams used their superior weapons to take control of the best grazing land. In the 1830s Jonker Afrikaner concluded an agreement with the Nama chief Oaseb whereby the Oorlams would protect the central grasslands of Namibia from the Herero who were then pushing south. In return Jonker Afrikaner was recognised as overlord, received tribute from the Nama, and settled at what today is Windhoek, on the borders of Herero territory. The Afrikaners soon came in conflict with the Herero who entered Damaraland from the south at about the same time as the Afrikaner started to expand farther north from Namaqualand. Both the Herero and the Afrikaner wanted to use the grasslands of Damaraland for their herds. This resulted in warfare between the Herero and the Oorlams as well as between the two of them and the Damara, who were the original inhabitants of the area. The Damara were displaced by the fighting and many were killed.

The it will be good to read a bloody page of the history about them:

https://en.m.wikipedia.org/wiki/Herero_and_Namaqua_genocide

shake well and thereafter get dates of old admixtures and even date the human race, even dancing over a dead cat would not make me more happy...

 

[berun]

OK Megalophias, the authors are aware of recent admixture:

All modern-day Khoe-San are drawn towards other 105 Africans and non-Africans compared to the ancient individuals from Ballito Bay, including Ju|’hoansi 106 San, thus far thought to be the least affected by recent admixture8, 9.

The 123 migration had a pronounced impact on all current Khoe-San groups, not only on the descendants of Stone 124 Age herders, such as the Khoekhoe (Fig. 2B, Extended Data Table 2, SI 6.4-6.5). This has been an elusive 125 result since all modern-day Khoe-San individuals display ≥9% recent admixture.

But the reference of Pickrell / Barbieri in "The genetic prehistory of southern Africa" is providing correct arguments for known migrations:

The positions of the Nama individuals in Supplementary Figure S2 are suggestive of post-colonial European admixture, in accordance with historic documentation of European ancestry in some Nama groups [23] . To test this, we used four-population tests [26] of the form [[Yoruba, X],[Han, French]], where X is any southern African population. A positive f4 statistic indicates gene ow between X and a population related to the French (or alternatively gene ow between populations related to the Yoruba and Han). The most strongly positive f4 statistic in the southern African populations is for the Nama (Supplementary Figure S6A), as expected if they have experienced European admixture. To conrm the direction of this gene ow, we used ROLLOFF [27] to test if there is detectable admixture LD in the Nama. If we use the Juj'hoan North and the French as the putative mixing populations, there is clear admixture LD (Supplementary Figure S6B), we date this mixture to approximately ve generations (150 years) ago.

They identify correctly a recent European ancestry (displayed in Figure S8 quite clearly: the Nama have around a 20% of Basque / French orange admixture), leaving so the Bantu ancestry as old...

These dates are consistent with archeological evidence for the arrival of both East African pastoralists and agriculturalists (probably Bantu speakers) in southern Africa 2,000–1,200 years ago32–35. PCA suggests that the majority of admixture in the Khoisan is more closely related to the Yoruba (from West Africa, linguistically related to Bantu speakers) than to the Dinka (from northeastern Africa; Supplementary Fig. S5), although our data are consistent with additional East African ancestry in some Khoe speakers (Supplementary Methods).

Instead, the recent paper is providing an old Eurasian/African phantom migration.

 

[holderlin]

hot damn. missed this one. Coinciding with the discovery of what initially appears to be AMH in North Africa 300k ya.

The picture will continue to get more interesting.

It continues to amaze me that every era of science thinks that "everything has been discovered".

 

[Megalophias]

OK Megalophias, the authors are aware of recent admixture....

Of course they are aware of recent admixture, for heaven's sake.

I have no idea why you think that cryptic colonial admixture that inexplicably appears many times older than it is and leaves no uniparental signature is a better explanation than old East African admixture that leaves the proper LD curve, an East African Y haplogroup, and East African lactase persistence alleles.

 

[berun]

They are aware of recent admixture... but they don't deal with it. In fact is just the contrary as *****ell-Barbieri paper as they blend recent European and old Bantu admixture as to get an Afroeurasian old admixture. For lactase persistence it is just was received from the first Bantu colonists, and it's shared with the Xhosa. For such east African haplo it's not shared by Bantu neighbours, right?

 

[Megalophias]

They are aware of recent admixture... but they don't deal with it. In fact is just the contrary as *****ell-Barbieri paper as they blend recent European and old Bantu admixture as to get an Afroeurasian old admixture. For lactase persistence it is just was received from the first Bantu colonists, and it's shared with the Xhosa. For such east African haplo it's not shared by Bantu neighbours, right?

There are plenty of papers on this subject. I find the East African scenario much more convincing than a magic hybrid Bantu + colonial Eurasian scenario. I'm heading out of town now, so you can have the last word.

 

[Angela]

A review of these recent papers by Razib Khan: The search for Eden
https://gnxp.nofe.me/2017/06/11/the...vistas/?utm_source=dlvr.it&utm_medium=twitter

 

[berun]

There are plenty of papers on this subject. I find the East African scenario much more convincing than a magic hybrid Bantu + colonial Eurasian scenario. I'm heading out of town now, so you can have the last word.

Do you take quantity as quality?

In whichever case even in the Khoisan tribes with less East African Eurasian admixture the genetists have found genetic "outliers", and that is like to say migrants... but everybody is free to believe in ghosts, or ghost pops.

 

[berun]

One of the many papers is that one written also by our beloved D. Reich "Ancient west Eurasian ancestry in southern and eastern Africa":

We began with an analysis of population mixture in southern Africa, using the data from Pickrell et al. (3) supplemented with
an additional 32 individuals from seven Khoisan populations genotyped on the Affymetrix Human Origins Array (SI Appendix,
Table S1); note that the Damara are excluded from most of the subsequent analyses because they genetically resemble southern
African Bantu-speaking groups (3). These southern African data were then combined with previously published worldwide
data (5) (SI Appendix). After removing individuals who seemed to be genetic outliers with respect to others in their population
(SI Appendix),

It's quite obvious the tricks here.

We thus computed weighted LD curves in the Juj’hoan_North, using the Juj’hoan_North themselves as one reference population and a range of 74 worldwide populations as the other, and examined the amplitudes of these curves (Fig. 1A). The largest amplitudes are obtained with European populations as references (Fig. 1A); taken literally, this would seem to implicate Europe as the source of admixture (although Middle Eastern populations are also among the best proxies). The estimated date for this gene flow is 43 ± 2 generations [1,290 ± 60 y, assuming 30 y per generation (7)] before the present, consistent with our previously estimated date (3). This date is well before the historical arrival of European colonists to the region.

big LOL!!

Under a model of admixture from a single source population, the decay rate of the LD curve does not depend on the
reference population used (6); this suggests that there are at least two separate non-Khoisan sources of ancestry in some of these
Khoisan populations.

yes! they have found two admixture events, maybe old Bantu and recent Coloured? let's see them...

The two inferred mean admixture times in the Gkana are 4 ± 1 and 39 ± 6 generations ago. ... By examining these amplitudes,
we conclude that the west Eurasian ancestry in the Gkana entered the population through the older admixture event (Fig. 3). Because of the caveat noted above, however, we cannot distinguish between two historical scenarios with this method: direct gene flow from a west Eurasian population and gene flow from a west Eurasian-admixed population.

superb!

The Taa_West also show two episodes of west Eurasian admixture, but the more recent one has low confidence. ... this method fails to detect that the Naro are admixed between two distinct Khoisan groups (3), and we find evidence of west African ancestry in just four Khoisan populations (!Xuun, Gkana, Khwe, and Taa_East) when treated individually (but see analyses of combined populations below).

a method also could fail if data is badly entered and interpreted:

in contrast, there is considerable genetic drift in west Eurasian populations because of the out-of-Africa bottleneck, which allows
admixture events to be more confidently assigned to this ancestry ... To increase our power to detect additional admixture
events, we performed analyses of combined populations. In a combined set of populations (the Tshwa, Shua, Haikom,
ǂHoan, Naro, and Taa_North) that have marginal evidence for a second, more recent admixture event, we infer two dates of
admixture: one 40 ± 2 generations ago and one 4 ± 1 generations ago (Z-score for the hypothesis that the admixture time is zero is
3.2, P = 7 × 10−4). In a combined set of two populations (the Juj’hoan_North and Gjui) that have marginal evidence for
a second, more ancient admixture event, we also infer two dates of admixture (SI Appendix, Fig. S24), but with different dates
from all other samples: one 30 ± 4 generations ago (Z-score of 6.9, P = 2 × 10−12) and one 109 ± 41 generations ago (Z-score of
2.6, P = 0:005). We interpret this as suggestive evidence that the population that introduced west Eurasian ancestry to southern
Africa was itself admixed, and that this more ancient admixture happened around 110 generations ago (although the confidence
intervals here are clearly large). ... The highest levels of west Eurasian ancestry are found in Khoe–Kwadi speakers (Table 1, southern Africa), particularly the Nama, where our estimate of west Eurasian ancestry reaches 14% (although note we cannot distinguish between the impact of recent colonialism and older west Eurasian ancestry in the Nama using this method).

∼1% of this is west Eurasian ancestry (Table 1) and the Juj’hoan_North have no Bantu-related ancestry, then this gives an admixture proportion of ∼25% west Eurasian ancestry in the putative eastern African source population. Using this value, we then estimated the proportions of Khoisan, putative eastern African, and Bantu-related ancestry of all populations using a linear model (18)

ops! another trick!

The table 8 of the supp is just telling what fiasco was done: the samples Bantu tribes have some 7-30% of Khoisan ancestry... but 0% of Eastern Africa/Eurasian ancestry... how in the hell it's possible that the Eurasian genes would be skipt when Bantus and Khoisans admixed in the huts? they had near the bed a DNA lab much more sophisticated than that of Reich?

By the way the paper is conscious about recent Eurasian admixture in the Nama, but is not saying nothing about genocide, rapes and camps in Namibia, or the existence of Coloureds.

 

[berun]

Also a pitfall to believe in a ghost migration of mixed East African / Eurasian population is the fact that Khoisans have 0% of haplo J, as it's shared by some 70% of Arabian Arabs and a 33% of Amhara (clealy the responsible to deliver the Semitic languages in Ethiopia). If the ghost migration was 25% Eurasian why Khoisans have no such haplo?

By the way some maps are quite expressive:

View attachment 8864

View attachment 8865

the presence of Afrikaans is debt to Oorlams and Basters from South-Africa, already mixed pops. In SA itself Afrikaans is the mother language of 60% of the whites, 76% of the Coloureds, 600000 Negroes and 60000 Asians.

View attachment 8866

this map reminds somehow the American "Indian Reservations"...

 

[berun]

This is from "Fine-scale human population structure in southern Africa reflects ecological boundaries":

This South / East Asian ancestry is not confined to the SAC [Couloured] population, as attested by the presence of the M36 mitochondrial haplogroup. The M36 mitochondrial haplogroup (South Indian/Dravidian in origin) is present in two out of 64 ≠Khomani San matrilineages,
(Table 1). The presence of M36 is likely derived from slaves of South Asian origin who escaped from Cape Town or the surrounding farms and dispersed into the northwestern region of South Africa. In addition, we observe one M7c3c lineage in the Nama (Table 1), which traces back to southeastern Asia but has been implicated in the Austronesian expansion of Polynesian speakers into Oceania (Kayser 2010; Delfin et al. 2012) and Madagascar (Poetsch et al. 2013). The importation of Malagasy slaves to Cape Town may best explain the observation of M7c3c in the Nama.

you can sum up the mutations of Indians, Austronesians, Khoesans, Dutchs, Xhosa, etc. and thereafter you get that the Homo sapiens is old as the dinosaurs... well, after looking how seem to dismiss data some genetists I will go to write down a paper like "Recent Corvus corax admixture in Europeans" as to be readily published in some respectful review as Nature or any yellow paper alike (they will sell more with such enthusiastic and awesome discoveries), and thereafter I will get more grants for my lab (institutions will verify that I publish in important pamphlets so I will be taken as a serious researcher). Red alarm: science had holy cows but now it has also charlatans.

 

[berun]

In Wood (2005) the Herero have a 8% I Y-DNA and 14% R..................................... I go to write down a paper about Homo sapiens dating about 1500000 years ago, and that the major diversity in Namibia of unrelated branches or more regional divergence provides evidence that Europeans are mainly pale Africans of recent arrival. Please deliver the grants to Bank of Bantustan before Reich or somelse would steal the idea.