PDA

View Full Version : Modern Nilotic Dinka wrong proxy for ancient African admixture in East Africans?



Mirix
02-02-2021, 12:25 PM
I was reading this one blog post about Ancient DNA in Ethiopia and i found what the author put down about how Modern Nilo-saharan speakers like the Dinka, Nuer, Shuluk and so on are wrong proxies for ancient African ancestry in East Africans interesting. Especially since we see so many on this forum emphasize Dinka related ancestry in East Africans and express it as ''Dinka like''

Thoughts on this?

African admixture component


When the Mota man was excavated in Ethiopia, some scholars wondered if the population to which he belonged might represent the native hunter-gatherers whom the Afro-Asiatic-speaking agropastoralists encountered and possibly mixed with when the latter first settled in the Horn. More extensive DNA analysis has, however, shown that gene flow between the two communities, although present, was minimal (Lipson et al. (2020 (https://www.nature.com/articles/s41586-020-1929-1)) places Mota-related admixture in the Cushitic-speaking Agaw at 8%; no other hunter-gatherer-related admixture has been detected in the Horn). Due to this apparent lack of a suitable ancient proxy sample, researchers have formulated alternative admixture scenarios using various modern populations. In qpAdm and other genomic programs, these contemporary reference groups have served as stand-ins for the elusive ancient African contact population(s). The Dinka, Nuer and other northern Nilotic communities have often been the go-to populations utilized as the African proxies. The problem is that all of these groups have proven to be even less reliable surrogates than Mota.

Let us examine why that is:

The Dinka, Nuer, Shilluk and other northern Nilotes are not ‘purely’ African populations to begin with. Like the Nilote groups to their south in the Horn (e.g. Kunama, Nara, Gumuz and Mursi) and Great Lakes (such as the Maasai, Kalenjin, Samburu and Turkana), the northern Nilotes also have considerable non-African admixture. This was already suggested by the uniparental lineages that they carry. Hassan et al. (2008) (http://khartoumspace.uofk.edu/bitstream/handle/123456789/17063/Y-Chromosome%20Variation%20Among%20Sudanese.pdf?sequ ence=1) observed that around 20% of Shilluk and 15% of Dinka individuals bear the E1b1b/E3b haplogroup, a paternal clade that is most common among Afro-Asiatic speakers. Most strikingly, Balemi (2018) (https://docdro.id/YdcUIHt) reports that 51.62% of his Nuer sample from southwestern Ethiopia carry the E1b1b-M78 haplogroup, compared to 16.67% of Hassan et al. (2008)’s Nuer sample from Sudan. Furthermore, 55.55% and 33.33% of Balemi (2018)’s Berta and Gumuz samples, respectively, bear the Eurasian F-M89 (https://en.wikipedia.org/wiki/Haplogroup_F-M89) clade. Non (2010) (http://etd.fcla.edu/UF/UFE0041981/non_a.pdf) likewise notes that over 40% of her Sudanese Nuer sample belongs to the mtDNA haplogroups M and N, including around 18% M1 carriers (cf. Table 3-3 (https://landofpunt.files.wordpress.com/2015/11/non2010.png)). The M1 (https://en.wikipedia.org/wiki/Haplogroup_M_(mtDNA)) subclade has been found in ancient Maghreban, Egyptian and South Cushitic specimens, and still remains a signature maternal lineage among the modern Afro-Asiatic-speaking populations in Northeast Africa. In short, uniparental markers indicate that there was significant gene flow from Afro-Asiatic-speaking groups (especially Cushitic speakers) into neighboring Nilote communities.
By contrast, in various autosomal DNA studies, the northern Nilotes have appeared to be almost completely of African ancestry. For example, in the African Genome Variation Project’s analysis, the Dinka sample showed no extraneous influence at K=2 (cf. Gurdasani et al. (2015) (https://www.repository.cam.ac.uk/bitstream/handle/1810/278857/nature13997.pdf)). This apparent lack of non-African affinities is why the Dinka, Nuer and other northern Nilotes have often been used by researchers as proxies for the African component. In 2017, Skoglund et al (https://www.cell.com/cell/pdf/S0092-8674(17)31008-5.pdf). compared the same AGVP Dinka sample to that of an ancient South Cushitic pastoralist (Luxmanda), the first such specimen to be genetically analyzed. In their admixture analysis, the Dinka individuals now all of a sudden showed almost 30% non-African ancestry at the K=2 level. Judging by the existing uniparental marker data, it’s pretty clear why that is: there was non-African ancestry buried within the northern Nilote gene pool, and that ancestry was specifically derived from earlier Cushitic peoples such as Luxmanda.

https://landofpunt.files.wordpress.com/2015/11/skoglund2017.png?w=290&h=600
Genome analysis detecting almost 30% non-African ancestry in a sample of Dinka. This Nilotic population was previously assumed to have little-to-no non-African ancestry, and was therefore often used as a proxy for inferring African ancestry (Skoglund et al. (2017 (https://www.cell.com/cell/pdf/S0092-8674(17)31008-5.pdf))).

This, in a nutshell, explains why the Dinka, Nuer and other northern Nilotes often fit well as proxies for the African component in admixture testing models. However, the problem with using such admixed groups in these analyses is that doing so leads to highly distorted estimates of ancestry proportions. For example, let’s say a researcher is testing admixture models on qpAdm and finds that his modern Cushitic or Ethiosemitic-speaking population is best modeled as 50% Neolithic Levantine + 50% Dinka. What that program is really telling him is that his sample is best modeled as 50% Neolithic Levantine + X% ancient Nilotic + X% ancient Cushitic. This again stems from the fact that the Dinka are not a purely African population, but rather a Nilote-descended community with significant Cushitic admixture (whence was derived their non-African ancestry). When factoring in the ~30% of non-African ancestry that Skoglund et al. (2017) detected in their Dinka reference sample, the estimated whole genome ancestry apportionment then actually becomes 50% Neolithic Levantine + 35% African + 15% unclassified Eurasian. This necessary adjustment therefore brings the non-African total to around 65%. The same corrective adjustment would have to be made if the African reference sample were a southern Nilotic group, such as the Maasai Nilotes (e.g. when the ~30% of West Eurasian admixture in Ali et al. (2020) (https://www.nature.com/articles/s41598-020-62645-0)‘s East African proxy sample (Maasai) is taken into account, the estimated non-African ancestry for their northeastern Somali sample rises to 70%; this more accurate total is close to Hodgson et al.’s 66% average for their general ethnic Somali sample). That’s also before correcting for linkage disequilibrium bias, which, as Hodgson et al. (2014) observed and Choudhury et al. (2020) (https://www.nature.com/articles/s41586-020-2859-7) also demonstrated, would increase the Cushitic speakers’ Eurasian total even further. Put simply, there is considerable extra non-African ancestry in the genome of the Afro-Asiatic-speaking sample which is not being counted. This results in an inaccurate overall estimation of ancestry proportions. It also conflicts with scientific data gathered through other means (viz. craniometric analysis, anthropometric analysis, linguistics).

Mirix
02-02-2021, 12:29 PM
https://landofpunt.files.wordpress.com/2015/11/wang2020-5.png?w=768&h=567
Principal Component Analysis indicates that the ancient Cushitic pastoralists of the Great Lakes and modern Afro-Asiatic-speaking populations of the Horn of Africa genetically cluster with Epipaleolithic Iberomaurusian specimens excavated at Taforalt (https://en.wikipedia.org/wiki/Taforalt) and early Neolithic individuals buried at Ifri n’Amr or Moussa (https://en.wikipedia.org/wiki/Ifri%20_N'Ammar), both situated in Morocco. These samples also plot near the Levant Chalcolithic cohort. This is consistent with Wang et al. (2020)’s admixture analysis and genome modeling, which both suggest that the oldest Cushitic settlers in East Africa were almost entirely of non-African ancestry (~90%) at the time of their arrival in the region (Wang et al. (2020) (https://advances.sciencemag.org/content/6/24/eaaz0183), Supplementary Material (https://advances.sciencemag.org/content/advances/suppl/2020/06/08/6.24.eaaz0183.DC1/aaz0183_SM.pdf)).



Genetic research has found that a significant portion of the core Sub-Saharan ancestry of the northern Nilotes (Dinka, Nuer, Shilluk, etc.) is related to West African populations (e.g. Skoglund et al. (2017)). However, Prendergast et al. (2008) remark that there was “no evidence of western African-related ancestry” in the ancient Cushitic individuals that they examined. Skoglund et al. (2017) and Wang et al. (2020) likewise did not detect any West African affinity in the early Cushitic specimens that they analyzed. This is hardly surprising since West African-related ancestry appears to have arrived in East Africa much later, during the Iron Age with the first Bantu settlers of the Great Lakes region. A specimen excavated at the Deloraine Farm in Kenya, the Rift Valley’s oldest agricultural site, is the earliest individual found to carry such West African-related ancestry (dated ca. 1170-970 BP; cf. Prendergast et al. (2008)). This makes modern Nilotic individuals unrealistic proxies for inferring an African admixture component among the ancient Cushitic settlers of the Pastoral Neolithic.

Keneki20
02-02-2021, 02:37 PM
Deleted.

gihanga.rwanda
02-02-2021, 02:48 PM
I actually agree with the premise that the Dinka and other Nilotic speakers aren’t a perfect stand in for the African ancestry in Cushitic speakers and related group, but not for the same reasons as the author of this blog post. This blog post sounds like it was written by the author of the “Land of Punt”, who seems to harbor some sort of agenda and doesn’t appear to have a very strong grasp on population genetics. How else could s/he take the cited ADMIXTURE analysis as evidence for substantial Eurasian gene-flow (30%) into Dinka? It’s as if African population substructure doesn’t exist.

We can use Dinka to model the African ancestry in Cushites because the two groups - Nilotic and Cushitic speakers respectively - probably share common descent from an ancient population - Ancient East African - that contributed to both groups separately. This doesn’t mean that we can equate Dinka with AEA. It’s been suspected for years that Nilotes have elevated West African-related ancestry, which is generally absent in Cushitic speakers. This suggests that the African ancestors of Cushitic speakers were related but distinct from Dinka and other Nilotic speakers in the present day.

That being said, I do suspect that Cushitic speakers have some direct Dinka-like ancestry, based on linguistic evidence and presence of certain uniparental lineages across NE Africa, but the genesis of Cushitic-related ancestry is probably a bit more complicated than population A + population B.

gihanga.rwanda
02-02-2021, 03:01 PM
Edit: As I suspected...

https://www.google.com/amp/s/landofpunt.wordpress.com/2015/11/09/ancient-dna-from-ethiopia/amp/

Mnemonics
02-02-2021, 05:15 PM
Dinka has too much Central African related admixture to be a direct proxy for the AEA ancestry in North Africa and the Horn but it's still mostly the same stuff.

beyoku
02-02-2021, 06:13 PM
The analysis is correct. The obvious bias just throws the whole thing left.
The E1b uniparental makers in Nilotics are not Eurasian.
Nilotes ARE a composite, but i would argue mostly a composite of North African and equatorial ancestry....in fact I personally would take it a step further and argue they are mostly North African as far as recent ancestry and migration patterns.

Once a genetic analysis of any African populations goes down that road that North Africans dont exist and there is a clean distinction between Eurasians and Sub Saharan Africans you might as well just discard it.
The same type of analysis used to show Nilotes as partially Eurasian or Ancient East Africa pastoralists predominantly Eurasian could be better put to use in showing how north African ancestry is ubiquitous among geographically African adjacent populations carrying African derived lineages of E, M,N, and U. IMO....This could have been good scholarship.

2 cents.

Angoliga
02-02-2021, 08:31 PM
Dinka has too much Central African related admixture to be a direct proxy for the AEA ancestry in North Africa and the Horn but it's still mostly the same stuff.

Discount any overt Central-African related ancestry, the bidirectional sahelian W-African admixture in Dinka (*exempt in Horners) would likely make it a more appropriate proxy for the bulk of AEA ancestry in North-(West) Africans.

Curious, what kind of "Central-African" related admixture are you referring to in Dinka: are you noticing KEN_LSA/Mota type affinities in qpadm?

Mirix
02-02-2021, 09:06 PM
I actually agree with the premise that the Dinka and other Nilotic speakers aren’t a perfect stand in for the African ancestry in Cushitic speakers and related group, but not for the same reasons as the author of this blog post. This blog post sounds like it was written by the author of the “Land of Punt”, who seems to harbor some sort of agenda and doesn’t appear to have a very strong grasp on population genetics. How else could s/he take the cited ADMIXTURE analysis as evidence for substantial Eurasian gene-flow (30%) into Dinka? It’s as if African population substructure doesn’t exist.

We can use Dinka to model the African ancestry in Cushites because the two groups - Nilotic and Cushitic speakers respectively - probably share common descent from an ancient population - Ancient East African - that contributed to both groups separately. This doesn’t mean that we can equate Dinka with AEA. It’s been suspected for years that Nilotes have elevated West African-related ancestry, which is generally absent in Cushitic speakers. This suggests that the African ancestors of Cushitic speakers were related but distinct from Dinka and other Nilotic speakers in the present day.

That being said, I do suspect that Cushitic speakers have some direct Dinka-like ancestry, based on linguistic evidence and presence of certain uniparental lineages across NE Africa, but the genesis of Cushitic-related ancestry is probably a bit more complicated than population A + population B.

The author argues that nilotic speakers have Ancient Cushitic admixture. According to him this is why they fit so well as a proxy.


In 2017, Skoglund et al. compared the same AGVP Dinka sample to that of an ancient South Cushitic pastoralist (Luxmanda), the first such specimen to be genetically analyzed. In their admixture analysis, the Dinka individuals now all of a sudden showed almost 30% non-African ancestry at the K=2 level. Judging by the existing uniparental marker data, it’s pretty clear why that is: there was non-African ancestry buried within the northern Nilote gene pool, and that ancestry was specifically derived from earlier Cushitic peoples such as Luxmanda.

But your explanation makes sense too , a related common yet distinct ancient descent can be an explanation for the absence of West African in Cushitic speakers and why they might share similar non African ancestry with ancient cushites. But on the flipside failing to take account for the non-African related ancestry in these populations don't they give you distorted African ancestry proportions?

Anyways i sent him a message to come debate and discuss his points, i think he probably has a decent grasp of genetics seems to have been writing about it for some time and but yeah he probably is ideologically driven which is why i made this thread asking for peoples in put and perspective on this. Don't particularly agree with many of his points and some of his terminological usages is outdated.

Mnemonics
02-02-2021, 09:17 PM
Discount any overt Central-African related ancestry, the bidirectional sahelian W-African admixture in Dinka (*exempt in Horners) would likely make it a more appropriate proxy for the bulk of AEA ancestry in North-(West) Africans.

Curious, what kind of "Central-African" related admixture are you referring to in Dinka: are you noticing KEN_LSA/Mota type affinities in qpadm?

The Z-scores for Shum Laka or Mbuti (when I use a modern right pop list)
seem to push plausible qpAdm models into the implausible range which indicates that Dinka has an excess of something present in both.

Mirix
05-08-2021, 05:05 PM
Turns out my initial post was correct after all. This study tested the hypothesis of African ancestry in HOA populations being derived from or not being differenitiated from the Nilo-Saharan(Dinkas) and came to the conclusion that the African ancestry in HOA are actually distinct. The African ancestry of HOA derive from a distinct African population.

https://journals.plos.org/plosgenetics/article?id=10.1371/journal.pgen.1004393

The hypothesis that African ancestry in the HOA is not distinct from that found in neighboring Nilo-Saharan speaking populations (hypothesis 1A above) requires a history of homogenizing inter-population migration or relatively recent common origin. In the case of homogenizing gene flow, a correlation between genetic and geographic distance might be expected, with nearby populations more alike than distant populations. We calculated within and between population gene identity (the probability that two randomly drawn alleles are identical by state) for all the populations included in the 4K partitioned dataset. We then used the between population gene identity estimates among the predominantly Nilo-Saharan ancestry Anuak, Gumuz, South Sudanese, and the African ancestry partition of the Amhara, Ari, Oromo, Somali, and Tygray to test for a relationship between genetic and geographic distance. No significant relationship was recovered (Mantel test, r = −0.28, p = 0.185) (Figure 3A).

Since the pattern of genetic variation in the African ancestry of Sudanese and HOA populations is not a good fit to one model of ongoing gene flow, we tested the hypothesis that there is population substructure within and between HOA and Nilo-Saharan populations using AMOVA [65] and hierarchical population tree models [66], [67]. First, within the HOA we used AMOVA to test for differentiation between linguistic groups – the Omotic speaking Ari, the Semitic speaking Amhara and Tygray, and the Cushitic speaking Oromo and Somali – and found a significant difference (ΦGT = 0.013, p<0.001). We also fit the HOA data to two population tree models, one without substructure and one with linguistically defined subgroups (Figure 3B), and found that the tree with the linguistic groups is a significantly better fit to the data (K = 30, df = 1, p = 4.3×10−8). Next, we tested for the presence of linguistically delineated subgroups within the Anuak, Gumuz, and South Sudanese. Most southern Sudanese populations speak languages in the Nilotic branch of the Nilo-Saharan language family and the Anuak language is also a Nilotic language [68]. The Gumuz language is either a highly divergent Nilo-Saharan language or a language isolate [69]. AMOVA reveals a statistically significant difference between these linguistic groups (ΦGT = 0.024, p<0.001) and the population tree with linguistically defined subgroups (Figure 3C) is a significantly better fit to the data than the tree without subgroups (K = 132, df = 1, p≈0). Finally, putting all of the populations together in an AMOVA analysis, we find significant differences between linguistic subgroups at both a macro level (Nilo-Saharan vs Afro-Asiatic) (ΦGT = 0.014, p<0.0001) and a micro level (Nilotic, Gumuz, Omotic, Semitic, Cushitic) (ΦGT = 0.022, p<0.0001). Population tree models with these groupings are a significantly better fit to the data than a tree without subgroups (K = 415 and 264, df = 1, p≈0) (Figure 3D). The tree with the larger subgroups (Nilo-Saharan vs Afro-Asiatic) is a slightly better fit to the data (Λ = 662) than the tree with the smaller subgroups (Λ = 812; smaller Λ values indicate better fit).

These results support the hypothesis from ADMIXTURE K≥11 of a distinct African ancestry with a long history in differentiated HOA populations (hypothesis 1B above) over the hypothesis from ADMIXTURE K≤10 that African ancestry in the HOA is not substantially differentiated from that found in neighboring populations (hypothesis 1A). In fact, our results suggest a rather more complicated history for these regional populations. Studies of further population samples from ethnic groups in and near the western and southern edges of the Ethiopian escarpment are sure to be interesting.

Mirix
05-12-2021, 11:47 PM
Forgot to add this from the study , as it is significant:


After partitioning the SNP data into African and non-African origin chromosome segments, we found support for a distinct African (Ethiopic) ancestry
and a distinct non-African (Ethio-Somali) ancestry in HOA populations. The African Ethiopic ancestry is tightly restricted to HOA populations and likely represents an autochthonous HOA population.


Because there is archaeological, historical, and linguistic evidence for contact with non-African populations beginning about 3,000 years ago, it has often been assumed that the non-African ancestry in HOA populations dates to this time. In this work, we find that the genetic composition of non-African ancestry in the HOA is distinct from the genetic composition of current populations in North Africa and the Middle East. With these data, we demonstrate that most non-African ancestry in the HOA cannot be the result of admixture within the last few thousand years, and that the majority of admixture probably occurred prior to the advent of agriculture.



The non-African ancestry in the HOA, which is primarily attributed to a novel Ethio-Somali inferred ancestry component, is significantly differentiated from all neighboring non-African ancestries in North Africa, the Levant, and Arabia. The Ethio-Somali ancestry is found in all admixed HOA ethnic groups, shows little inter-individual variance within these ethnic groups, is estimated to have diverged from all other non-African ancestries by at least 23 ka, and does not carry the unique Arabian lactase persistence allele that arose about 4 ka. Taking into account published mitochondrial, Y chromosome, paleoclimate, and archaeological data, we find that the time of the Ethio-Somali back-to-Africa migration is most likely pre-agricultural.
http://journals.plos.org/plosgenetics/article?id=10.1371/journal.pgen.1004393

All the "Dinka Like" related claims are out the water it seems. Both the Non-African and African ancestry in the HOA population is distinct and tightly restricted to the Horn of African populations heavily differentiated from all other neighboring African and Non-African populations and African ancestry represents an autochtonous(indigenous) HOA population . And the study accounted this for by homogenizing gene flow. This significantly removes skewed admixture results and formation of false proxies that created the Dinka/Nilote associations to begin with. Which i related about in my first two posts.

The last part about the non-African ancestry being pre-agricultutre accords well with the linguistic and archeological evidence of the antiquity of Agriculture in HOA and Ehret shows how it owes nothing to Arabian introduction or influence. . On the Antiquity of Agriculture in Ethiopia (https://www.jstor.org/stable/181512?seq=3#metadata_info_tab_contents). It was often assumed by people that agriculture in the Horn of Africa owes much to South Arabian influence and migrations.

Mnemonics
05-13-2021, 08:30 AM
Forgot to add this from the study , as it is significant:





http://journals.plos.org/plosgenetics/article?id=10.1371/journal.pgen.1004393

All the "Dinka Like" related claims are out the water it seems. Both the Non-African and African ancestry in the HOA population is distinct and tightly restricted to the Horn of African populations heavily differentiated from all other neighboring African and Non-African populations and African ancestry represents an autochtonous(indigenous) HOA population . And the study accounted this for by homogenizing gene flow. This significantly removes skewed admixture results and formation of false proxies that created the Dinka/Nilote associations to begin with. Which i related about in my first two posts.

The last part about the non-African ancestry being pre-agricultutre accords well with the linguistic and archeological evidence of the antiquity of Agriculture in HOA and Ehret shows how it owes nothing to Arabian introduction or influence. . On the Antiquity of Agriculture in Ethiopia (https://www.jstor.org/stable/181512?seq=3#metadata_info_tab_contents). It was often assumed by people that agriculture in the Horn of Africa owes much to South Arabian influence and migrations.

The distinct Ethiopic ancestry they're talking about in that study is Mota-like ancestry which is incredibly diluted in most Horner Cushitic and Ethio-Semitc speakers who are all predominantly Dinka-like with the exception of the Oromo who have nearly equal amounts of Mota and Dinka ancestry (likely due to admixture from Omotic speakers)

Nevertheless Mota-like and Nilotic populations all seem to be deeply related with Mota having deeper African and OOA affinities.

From what I've seen it seems like the Horn has become progressive more Dinka-like over time with the specific Early_Pastoral Neolithic sample that seems to be ancestral to most of the following populations serving as the starting point.

The Dinka themselves also seem to have received extra Central and/or West African ancestry which makes them imperfect fits for the Dinka-like ancestry in Horners.

Mirix
05-13-2021, 03:35 PM
The distinct Ethiopic ancestry they're talking about in that study is Mota-like ancestry which is incredibly diluted in most Horner Cushitic and Ethio-Semitc speakers who are all predominantly Dinka-like with the exception of the Oromo who have nearly equal amounts of Mota and Dinka ancestry (likely due to admixture from Omotic speakers)

Nevertheless Mota-like and Nilotic populations all seem to be deeply related with Mota having deeper African and OOA affinities.

From what I've seen it seems like the Horn has become progressive more Dinka-like over time with the specific Early_Pastoral Neolithic sample that seems to be ancestral to most of the following populations serving as the starting point.

The Dinka themselves also seem to have received extra Central and/or West African ancestry which makes them imperfect fits for the Dinka-like ancestry in Horners.

My point was their admixture have skewered results which has led people to formulate admixture hypothesis using modern populations some people have tried to hypothesize that it is from an Nilotic Dinka sources (Which people in this forum accept without question) and subsequently used it as a proxy. In study i have shown above they divided the Ethio-Somali samples from all other African population samples. and they divided/groupe the HOA DNA into African and Non-African. The African DNA they found to be distinct and highly differentiated . They tested it against the Dinka hypothesis that like i explained on the first page and in the 2nd post above you


It confirms the reason why Dinka populations scores well it's because they are actually admixed populations themselves from cushites(who introduced in them non-African related ancestry) and not purely African and that just gives you distorted African ancestry proportions that is not accounted for. It may not actually show an actual relationship at all but a percieved one created by failing to take account of that. When you do take that into account by homogenizing the gene flow or control for relatively recent origin it removes that distortion so that you can get a clearer picture. What they found was no significant relationship. The African ancestry in HOA highly differentiate and distinct from the neighboring Nilo-Saharan speaking populations.


The Early Pastoral Neolithic could be good candidate for the source of West Euroasian ancestry for HOA i must say.

Mnemonics
05-13-2021, 07:04 PM
My point was their admixture have skewered results which has led people to formulate admixture hypothesis using modern populations some people have tried to hypothesize that it is from an Nilotic Dinka sources (Which people in this forum accept without question) and subsequently used it as a proxy. In study i have shown above they divided the Ethio-Somali samples from all other African population samples. and they divided/groupe the HOA DNA into African and Non-African. The African DNA they found to be distinct and highly differentiated . They tested it against the Dinka hypothesis that like i explained on the first page and in the 2nd post above you


It confirms the reason why Dinka populations scores well it's because they are actually admixed populations themselves from cushites(who introduced in them non-African related ancestry) and not purely African and that just gives you distorted African ancestry proportions that is not accounted for. It may not actually show an actual relationship at all but a percieved one created by failing to take account of that. When you do take that into account by homogenizing the gene flow or control for relatively recent origin it removes that distortion so that you can get a clearer picture. What they found was no significant relationship. The African ancestry in HOA highly differentiate and distinct from the neighboring Nilo-Saharan speaking populations.


The Early Pastoral Neolithic could be good candidate for the source of West Euroasian ancestry for HOA i must say.

The Dinka are very unlikely to be admixed with Cushitic populations. They seem to be less Neanderthal than the Yoruba and they lack a strong preference for West Eurasians over East Eurasians, which would be present if they were Cushitic admixed.

Regardless of whether or not the Dinka themselves contributed to Horn populations it is very evident that Horner populations have become more Dinka-like overtime which is impossible if you believe that this affinity is because of admixture from Cushitic populations to the Dinka.

The Early pastoralist who seems to be ancestral to the latter PN populations has rather minor Dinka affinity (8-15%) and is one of the most Natufian samples we have period.

The later pastoralists on the other had are mostly an even split between Dinka-like and Mota-like ancestry ( excluding the less Eurasian ones who are obviously admixed with additional Mota-like hunter gatherer ancestry)

Modern Horners are even more Dinka-like with the Amhara, Tigray, and Ethiopian Somalis showing high to medium single digits of Mota ancestry and the Somalis from Somalia proper showing even less.

This pattern wouldn't be present if we assume that any affinity between the populations was from Cushitic admixture into the Dinka.

Mirix
05-15-2021, 04:49 PM
The Dinka are very unlikely to be admixed with Cushitic populations. They seem to be less Neanderthal than the Yoruba and they lack a strong preference for West Eurasians over East Eurasians, which would be present if they were Cushitic admixed.

Regardless of whether or not the Dinka themselves contributed to Horn populations it is very evident that Horner populations have become more Dinka-like overtime which is impossible if you believe that this affinity is because of admixture from Cushitic populations to the Dinka.

The Early pastoralist who seems to be ancestral to the latter PN populations has rather minor Dinka affinity (8-15%) and is one of the most Natufian samples we have period.

The later pastoralists on the other had are mostly an even split between Dinka-like and Mota-like ancestry ( excluding the less Eurasian ones who are obviously admixed with additional Mota-like hunter gatherer ancestry)

Modern Horners are even more Dinka-like with the Amhara, Tigray, and Ethiopian Somalis showing high to medium single digits of Mota ancestry and the Somalis from Somalia proper showing even less.

This pattern wouldn't be present if we assume that any affinity between the populations was from Cushitic admixture into the Dinka.

What do you mean unlikely to be mixed? Have you actually read my opening post? And its not unlikely at all when they are neighboring population to Afro-Asiatic speakers in the nile valley they could very well have experienced gene flow from.

Several studies do in fact show that they do carry West Euro Asian paternal haplogroups common to cushitic/Afroc-Asiatic speaking populations. For example in the study on Restricted Gene flow among the Sudanese made by Hassan et al. (2008) (http://khartoumspace.uofk.edu/bitstream/handle/123456789/17063/Y-Chromosome%20Variation%20Among%20Sudanese.pdf?sequ ence=1) observed that around 20% of Shilluk and 15% of Dinka individuals bear the E1b1b/E3b haplogroup, a paternal clade that is most common among Afro-Asiatic speakers. Most strikingly, Balemi (2018) (https://www.docdroid.net/YdcUIHt/genetic-study-of-lct-enhancer-y-chromosomeand-mitochondrial-dna-variation-in-some-ethnic-groups-in-ethiopia-adugna-2018-pdf) reports that 51.62% of his Nuer sample from southwestern Ethiopia carry the E1b1b-M78 haplogroup, compared to 16.67% of Hassan et al.

And then another Genome analysis detecting almost 30% non-African ancestry in a sample of Dinka. This Nilotic population was previously assumed to have little-to-no non-African ancestry, and was therefore often used as a proxy for inferring African ancestry (Skoglund et al. (2017)) (https://www.cell.com/cell/pdf/S0092-8674(17)31008-5.pdf)

Can you define what Dinka Like actually means? because you just repeat it over and over . The African Ancestry in Horn of Africans don't bear any relationship with the neighboring Nilo-Saharan populations as the study i qouted above you shows and when taken with homogenizing of gene flow and controlling for recent common origin they found that it was actually distinct from it. So the question remains how would they be Dinka like ?

''Dinka like" sounds like just an assumption being made here considering the above evidence shows Nilo-Saharan populations being considerably admixed with AA populations.

Mnemonics
05-15-2021, 08:35 PM
What do you mean unlikely to be mixed? Have you actually read my opening post? And its not unlikely at all when they are neighboring population to Afro-Asiatic speakers in the nile valley they could very well have experienced gene flow from.

Several studies do in fact show that they do carry West Euro Asian paternal haplogroups common to cushitic/Afroc-Asiatic speaking populations. For example in the study on Restricted Gene flow among the Sudanese made by Hassan et al. (2008) (http://khartoumspace.uofk.edu/bitstream/handle/123456789/17063/Y-Chromosome%20Variation%20Among%20Sudanese.pdf?sequ ence=1) observed that around 20% of Shilluk and 15% of Dinka individuals bear the E1b1b/E3b haplogroup, a paternal clade that is most common among Afro-Asiatic speakers. Most strikingly, Balemi (2018) (https://www.docdroid.net/YdcUIHt/genetic-study-of-lct-enhancer-y-chromosomeand-mitochondrial-dna-variation-in-some-ethnic-groups-in-ethiopia-adugna-2018-pdf) reports that 51.62% of his Nuer sample from southwestern Ethiopia carry the E1b1b-M78 haplogroup, compared to 16.67% of Hassan et al.

And then another Genome analysis detecting almost 30% non-African ancestry in a sample of Dinka. This Nilotic population was previously assumed to have little-to-no non-African ancestry, and was therefore often used as a proxy for inferring African ancestry (Skoglund et al. (2017)) (https://www.cell.com/cell/pdf/S0092-8674(17)31008-5.pdf)

Can you define what Dinka Like actually means? because you just repeat it over and over . The African Ancestry in Horn of Africans don't bear any relationship with the neighboring Nilo-Saharan populations as the study i qouted above you shows and when taken with homogenizing of gene flow and controlling for recent common origin they found that it was actually distinct from it. So the question remains how would they be Dinka like ?

''Dinka like" sounds like just an assumption being made here considering the above evidence shows Nilo-Saharan populations being considerably admixed with AA populations.

I suppose I should have been more clear. The Dinka are unlikely to have "significant" admixture from the Horn. The Gumuz are an excellent example of Nilotic population that have minor but significant admixture from Horner-like populations.

When I'm talking about Dinka-like ancestry I am referring to the specific strain of AEA ancestry that peaks in Nilotic populations. It is very distinct from the native Mota-like ancestry found in the Horn dispite both populations being likely very related.

Let us assume that you are correct and that the Dinka have large amounts of Horner ancestry, you have to answer three very important questions.

Why do they lack the strong preference for West Eurasians present in Horners?

This is important because if they were significantly admixed with West Eurasian ancestry they should strongly prefer Ancient West Eurasians strongly over East Eurasians. You could get around that by assigning a special relationship between their non-Horner admixture and East Eurasians but that is not a particularly parsimonious explanation.

Why do they have slightly less Neanderthal than Yoruba?

Neanderthal can act as a sort of tracer dye for Eurasian ancestry which allows us to determine if a population has significant Crown Eurasian ancestry. Horners obviously have Neanderthal ancestry proportional to their Eurasian admixture. The Dinka on the other hand seem to have a lower affinity to Neanderthals as compared to the Yoruba.

If we assume that the African ancestors of the Dinka got all of their Neanderthal from Horner populations (which is incredibly unlikely considering they have a minor affinity to Iberomaurusians) we are looking at a peak of about 10-20% Horner admixture which would not explain the strength of the admixture models and has also some how left no strong affinity for Ancient West Eurasians.

Why does the affinity to Dinka increase in the Horn?

Populations in the Horn seem to have gotten more Dinka-like overtime with Early_PN followed by Kenya_PastoralN and then Kenya_PastoralIA, each of which is more Dinka-like than the last.

How is this occurring if admixture did not occur?
We already know that Mota like populations were native to the Horn and the existence of the Southeast African cline confirms that this population was relatively old in the region. Which means that the the various pastoralists populations got successively less native overtime.

Is that possible without admixture from an external source?


The presence of uniparentals like E1b1b in Nilotic populations are pretty easily explained by founder effects after minor admixture from neighboring pastoralists. But, everything else cannot be explained by your thesis.

Riverman
05-15-2021, 11:19 PM
I suppose I should have been more clear. The Dinka are unlikely to have "significant" admixture from the Horn. The Gumuz are an excellent example of Nilotic population that have minor but significant admixture from Horner-like populations.

When I'm talking about Dinka-like ancestry I am referring to the specific strain of AEA ancestry that peaks in Nilotic populations. It is very distinct from the native Mota-like ancestry found in the Horn dispite both populations being likely very related.

Let us assume that you are correct and that the Dinka have large amounts of Horner ancestry, you have to answer three very important questions.

Why do they lack the strong preference for West Eurasians present in Horners?

This is important because if they were significantly admixed with West Eurasian ancestry they should strongly prefer Ancient West Eurasians strongly over East Eurasians. You could get around that by assigning a special relationship between their non-Horner admixture and East Eurasians but that is not a particularly parsimonious explanation.

Why do they have slightly less Neanderthal than Yoruba?

Neanderthal can act as a sort of tracer dye for Eurasian ancestry which allows us to determine if a population has significant Crown Eurasian ancestry. Horners obviously have Neanderthal ancestry proportional to their Eurasian admixture. The Dinka on the other hand seem to have a lower affinity to Neanderthals as compared to the Yoruba.

If we assume that the African ancestors of the Dinka got all of their Neanderthal from Horner populations (which is incredibly unlikely considering they have a minor affinity to Iberomaurusians) we are looking at a peak of about 10-20% Horner admixture which would not explain the strength of the admixture models and has also some how left no strong affinity for Ancient West Eurasians.

Why does the affinity to Dinka increase in the Horn?

Populations in the Horn seem to have gotten more Dinka-like overtime with Early_PN followed by Kenya_PastoralN and then Kenya_PastoralIA, each of which is more Dinka-like than the last.

How is this occurring if admixture did not occur?
We already know that Mota like populations were native to the Horn and the existence of the Southeast African cline confirms that this population was relatively old in the region. Which means that the the various pastoralists populations got successively less native overtime.

Is that possible without admixture from an external source?


The presence of uniparentals like E1b1b in Nilotic populations are pretty easily explained by founder effects after minor admixture from neighboring pastoralists. But, everything else cannot be explained by your thesis.

I think these are good points, but wouldn't most of it be answered by the alternative of the Eurasian admixture being mostly (not exclusively) Basal Eurasian derived? This would also make E1b1b easier to explain. How would an admixture model with a population which is much more Basal Eurasian shifted than Natufians fare?

Mnemonics
05-16-2021, 02:29 AM
I think these are good points, but wouldn't most of it be answered by the alternative of the Eurasian admixture being mostly (not exclusively) Basal Eurasian derived? This would also make E1b1b easier to explain. How would an admixture model with a population which is much more Basal Eurasian shifted than Natufians fare?

His claim is that the Dinka are admixed with populations from the Horn, which would indicate that they are significantly West Eurasian and/or significant Mota like. Both of which are made unlikely by those factors.

AEA is likely an admixed component itself but I doubt that is the result of recent admixture.

I'm not even sure that "Basal Eurasian" proper exists anymore. The assumption that it exists is due to the inability of fstats to find significant relationships between Basal Eurasian rich populations and SSA populations but there is an obvious issue with the ability of the software to detect significant relationships between SSA populations and heavily admixed AEA rich populations like Kenya_PastoralN.

It gets so bad that you can put two members of the same admixed African population in the fstat and get large significant stats for the Eurasian population even when you switch them.

Mirix
05-16-2021, 08:44 PM
I suppose I should have been more clear. The Dinka are unlikely to have "significant" admixture from the Horn. The Gumuz are an excellent example of Nilotic population that have minor but significant admixture from Horner-like populations.
When I'm talking about Dinka-like ancestry I am referring to the specific strain of AEA ancestry that peaks in Nilotic populations. It is very distinct from the native Mota-like ancestry found in the Horn dispite both populations being likely very related.

Actually Gumuz according to Balemi (2018) (https://www.docdroid.net/YdcUIHt/genetic-study-of-lct-enhancer-y-chromosomeand-mitochondrial-dna-variation-in-some-ethnic-groups-in-ethiopia-adugna-2018-pdf) 55.55% and 33.33% of Balemi (2018)’s Berta and Gumuz samples, respectively, bear the Eurasian F-M89 clade.

But it is not just Gumuz or Dinka , for example Non (2010) (http://etd.fcla.edu/UF/UFE0041981/non_a.pdf) likewise notes that over 40% of her Sudanese Nuer sample belongs to the mtDNA haplogroups M and N, including around 18% M1 carriers (cf. Table 3-3). The M1 subclade has been found in ancient Maghreban, Egyptian and South Cushitic specimens, and still remains a signature maternal lineage among the modern Afro-Asiatic-speaking populations in Northeast Africa.

In short, uniparental markers indicate that there was significant gene flow from Afro-Asiatic-speaking groups (especially Cushitic speakers) into neighboring Nilote communities.

Can you expand on the specific AEA ancestry?


Let us assume that you are correct and that the Dinka have large amounts of Horner ancestry, you have to answer three very important questions.

Why do they lack the strong preference for West Eurasians present in Horners? This is important because if they were significantly admixed with West Eurasian ancestry they should strongly prefer Ancient West Eurasians strongly over East Eurasians. You could get around that by assigning a special relationship between their non-Horner admixture and East Eurasians but that is not a particularly parsimonious explanation.
Why do they have slightly less Neanderthal than Yoruba? Neanderthal can act as a sort of tracer dye for Eurasian ancestry which allows us to determine if a population has significant Crown Eurasian ancestry. Horners obviously have Neanderthal ancestry proportional to their Eurasian admixture. The Dinka on the other hand seem to have a lower affinity to Neanderthals as compared to the Yoruba. If we assume that the African ancestors of the Dinka got all of their Neanderthal from Horner populations (which is incredibly unlikely considering they have a minor affinity to Iberomaurusians) we are looking at a peak of about 10-20% Horner admixture which would not explain the strength of the admixture models and has also some how left no strong affinity for Ancient West Eurasians.

Aren't these assumptions being made here on your part?. Because we are extrapulating on populations that was assumed to be less admixed then before. For example in various autosomal DNA studies, the northern Nilotes have appeared to be almost completely of African ancestry. For example, in the African Genome Variation Project’s analysis, the Dinka sample showed no extraneous influence at K=2 (cf. Gurdasani et al. (2015 (https://www.repository.cam.ac.uk/bitstream/handle/1810/278857/nature13997.pdf)))

But on the studies i have shown above that proves to not be the case.



Why does the affinity to Dinka increase in the Horn?

Populations in the Horn seem to have gotten more Dinka-like overtime with Early_PN followed by Kenya_PastoralN and then Kenya_PastoralIA, each of which is more Dinka-like than the last.

How is this occurring if admixture did not occur?
We already know that Mota like populations were native to the Horn and the existence of the Southeast African cline confirms that this population was relatively old in the region. Which means that the the various pastoralists populations got successively less native overtime.

Is that possible without admixture from an external source?


The presence of uniparentals like E1b1b in Nilotic populations are pretty easily explained by founder effects after minor admixture from neighboring pastoralists. But, everything else cannot be explained by your thesis.


The presence of uniparentals are more likely to have come ancient Cushites . In 2017, Skoglund et al. (https://www.cell.com/cell/pdf/S0092-8674(17)31008-5.pdf) compared the same AGVP Dinka sample to that of an ancient South Cushitic pastoralist (Luxmanda), the first such specimen to be genetically analyzed.

In their admixture analysis, the Dinka individuals now all of a sudden showed almost 30% non-African ancestry at the K=2 level. Judging by the existing uniparental marker data, it’s pretty clear why that is: there was non-African ancestry buried within the northern Nilote gene pool, and that ancestry was specifically derived from earlier Cushitic peoples such as Luxmanda.

If there is any percieved affinity it could easily be explained when taking the above information in. This taking to together with the fact the ancient Cushitic specimens are devoid of West African affinity , makes modern Nilotic individuals unrealistic proxies for inferring an African admixture component among the ancient Cushitic settlers of the Pastoral Neolithic.

Mirix
05-16-2021, 08:57 PM
His claim is that the Dinka are admixed with populations from the Horn, which would indicate that they are significantly West Eurasian and/or significant Mota like. Both of which are made unlikely by those factors.
AEA is likely an admixed component itself but I doubt that is the result of recent admixture.

I'm not even sure that "Basal Eurasian" proper exists anymore. The assumption that it exists is due to the inability of fstats to find significant relationships between Basal Eurasian rich populations and SSA populations but there is an obvious issue with the ability of the software to detect significant relationships between SSA populations and heavily admixed AEA rich populations like Kenya_PastoralN.

It gets so bad that you can put two members of the same admixed African population in the fstat and get large significant stats for the Eurasian population even when you switch them.

My claim is that the neighboring Nilotic populations are not purely African or to less admixed as it was previously thought and have non-African ancestry buried in their gene pool. They show obvious gene flow from ancient AA speaking groups (Especially ancient cushitic populations) when you look at their uniparental markers. This not being taking into account has created distorted and skewed ancestry proportions.


What also makes me doubt this supposed affinity is that Wang et al. (2020) (https://advances.sciencemag.org/content/6/24/eaaz0183)’s admixture analysis and genome modeling, which both suggest that the oldest Cushitic settlers in East Africa were almost entirely of non-African ancestry (~90%) at the time of their arrival in the region.

Also the simple fact that they do have West African Affinity and the Early Neolothic Pastoral specimens are completely devoid of it , makes them unrealistic/unfit proxies for inferring an African admixture component among the ancient Cushitic settlers of the Pastoral Neolithic.

Mnemonics
05-16-2021, 10:54 PM
My claim is that the neighboring Nilotic populations are not purely African or to less admixed as it was previously thought and have non-African ancestry buried in their gene pool. They show obvious gene flow from ancient AA speaking groups (Especially ancient cushitic populations) when you look at their uniparental markers. This not being taking into account has created distorted and skewed ancestry proportions.


What also makes me doubt this supposed affinity is that Wang et al. (2020) (https://advances.sciencemag.org/content/6/24/eaaz0183)’s admixture analysis and genome modeling, which both suggest that the oldest Cushitic settlers in East Africa were almost entirely of non-African ancestry (~90%) at the time of their arrival in the region.

Also the simple fact that they do have West African Affinity and the Early Neolothic Pastoral specimens are completely devoid of it , makes them unrealistic/unfit proxies for inferring an African admixture component among the ancient Cushitic settlers of the Pastoral Neolithic.

The most Eurasian Early Pastoralist's SSA is predominantly Mota-related which indicates that the initial population was heavily Eurasian. But, their non-Native SSA is entirely Dinka-like which indicates that the population received Dinka-like admixture before arriving in the Horn which doesn't support your idea that the Dinka received significant admixture from Kenya_PastoralN type populations.

Luxmanda is an interesting sample in that it seems to be a mix of Early Pastoralist + Mota which kind of resembles modern Omotic speakers. Its SSA is predominantly native to the Horn and distinct from Dinka-like ancestry.

Chad Rohlfsen
05-16-2021, 11:08 PM
Direct admixture from Shum Laka to the Dinka is something that does pop up in qpGraph.

Mirix
05-17-2021, 10:45 AM
The most Eurasian Early Pastoralist's SSA is predominantly Mota-related which indicates that the initial population was heavily Eurasian. But, their non-Native SSA is entirely Dinka-like which indicates that the population received Dinka-like admixture before arriving in the Horn which doesn't support your idea that the Dinka received significant admixture from Kenya_PastoralN type populations.

Luxmanda is an interesting sample in that it seems to be a mix of Early Pastoralist + Mota which kind of resembles modern Omotic speakers. Its SSA is predominantly native to the Horn and distinct from Dinka-like ancestry.

The Mota fossils are not representative of the Early ancestral Afro-Asiatic speakers in the Horn. Lorente et al. note that contemporary Ethiopian populations (specifically, the Afro-Asiatic speakers) have substantial West Eurasian affinities that the Mota specimen does not appear to possess.

And as i explained earlier the Earliest Cushitic settlers appear to have been almost entirely non-African ancestry at the time. Only around 1,300 years after the earliest radiocarbon-dated Pastoral Neolithic sample does the first probable evidence of Nilotic influence in form of the Y-DNA haplogroup A and the mtDNA clade L4, appear among specimens excavated at the Naishi Rockshelter and Keringet Cave.

The comparison between the ancient South Cushitic pastoralist (Luxmanda) and the Dinka sample shows obvious gene flow into them, beceause the Dinka sample all of a sudden showed almost 30% non-African ancestry at the K=2 level. This together with their various significant uniparental markers of mtDNA haplogroups M and N, and, E1b1b/E3b haplogroup, and M78, F-M89 clades shows they particularly admixed with ancient cushitic speakers.
https://anthrogenica.com/showthread.php?22880-Modern-Nilotic-Dinka-wrong-proxy-for-ancient-African-admixture-in-East-Africans&p=771804#post771804

This in a nutshell is why they seemingly fit so well as a proxy.

CopperAxe
05-18-2022, 01:31 PM
I'm a self admitted noob when it comes to deep ancestry in Africa, primarily because the lack of ancient samples. But if the (non-admixed) Nilotes are hypothetically a mix between predominantly AEA with input from a Central/Sahelian African population, would it be worthwhile to make ghost coordinates using several Nilotic populations?

For example Dinka minus x% west African ancestry = AEA, and Dinka minus x% Mota(?) = Central/Sahelian population.

I guess two issues here would be that A; This central African population probably wouldn't have the same degrees of basal African ancestry that West Africans have and B; Mota might have some slight eurasian ancestry.

What is the best proxy for the basal African in west Africans anyways? I generally use the Aka/Biaka coordinates because you get somewhat realistic proportions and it shows up only with people who have ancestry from west Africa. It is also the closest to Shum Laka iirc.

Thoughts would be appreciated.