FAQ

1. What is this website about?
2. Where was the ancient Land of Punt?
3. Who were the Puntites?
4. Why did the ancient Egyptians refer to the Puntites as brbrta if they were Cushitic speakers rather than Berber speakers?
5. What evidence is there that the Puntites spoke an Afro-Asiatic language?
6. What is the etymology of the term “Punt”?
7. Could the ancient “Puntite” statues that were excavated in Somalia actually be forgeries?
8. Were the ancient Egyptians related to the Puntites?
9. What is the genetic composition of the modern Afro-Asiatic-speaking populations of the Horn of Africa?
10. Do the ancient Cushites share this ancestral makeup?
11. Do modern South Omotic speakers share this ancestral makeup?
12. How come there aren’t any Northern Somali, Saho, Tuareg Berber or Nubian samples in your Vahaduo genome analysis?
13. If the Cushitic, Ethiosemitic and North Omotic-speaking populations of the Horn really do have a predominant non-African ancestry, why do some analyses claim otherwise?
14. Can Cushitic, Ethiosemitic and North Omotic-speaking individuals be genetically modelled as a simple two-way admixture of West Eurasians and Sub-Saharan Africans?
15. Can the ancestral composition of the Cushitic, Ethiosemitic and North Omotic-speaking populations realistically be modelled in qpAdm with a five-way admixture scenario?
16. If the Cushitic, Ethiosemitic and North Omotic-speaking populations of the Horn really do have a predominant non-African ancestry, why is this not more clearly reflected on a PCA map?
17. Were the ancient Azanians Cushites?
18. If Muslim Egyptians have other admixture elements besides the ancient Egyptian component, does this mean they are mostly of non-Egyptian origin?
19. You indicate that you used the Nakht-Ankh cohort to arrive at your conclusion that Afro-Asiatic speakers from the Horn and North Africa carry some European Steppe and East Asian admixture. Since Nakht-Ankh is not an official Global25 sample, could it be that that is the actual reason behind this result?
20. Is it true that the ancient Egyptians, like the Minoans, descend from Anatolian Neolithic farmers?
21. Did the Capsians of North Africa descend from the Iberomaurusians?
22. What are the genetic affinities of the Aterians?
23. Are you sure that the oblique eye folds, which are common among Afro-Asiatic speakers and Nubians in the Horn and North Africa, are not due to absorption of Khoisan peoples? What makes you ascribe this trait to Central Asian or East Asian admixture?
24. Is there any other evidence of an ancient Central Asian or East Asian-related presence in the Nile Valley?
25. What complexion were the ancient Cushitic settlers of eastern Africa?
26. Were the ancient Cushites slender?
27. Did the ancient Cushites have soft-textured hair?
28. Have the Kiffians and Tenerians of the Gobero been genetically analysed yet?
29. Is it true that the ancient Egyptian Pharaoh Ramesses III and the Prince Pentawere (Unknown Man E) belong to the E1b1a haplogroup?
30. Is it true that the ancient Egyptian Amarna royal family shares closest genetic affinities with Sub-Saharan African populations?
31. In your autosomal DNA analysis of the ancient Egyptian Amarna royal family, could the South Asian affinity you detected actually be related to the Andamanese Hunter-Gatherers (Ancestral South Indians) instead of European Steppe populations?
32. Which ancient DNA studies do you think should be published?
33. Who were the ancient Nubians?
34. If Nubians are ancestrally related to Egyptians, why do they now speak a Nilo-Saharan language instead of an Afro-Asiatic language?
35. Are Sudanese “Arabs” really peninsular Arabs?
36. Are Maghrebi “Arabs” really peninsular Arabs?
37. Are Ashkenazi Jews and Sephardic Jews really of the same ancestral origin as Mizrahi Jews, Palestinians, Lebanese, peninsular Arabs, Assyrians, Mahra and other Semites?
38. Which commercial genetic test, if any, do you recommend?
39. I am an Amhara from Ethiopia. Why are my 23andme and AncestryDNA test results different from each other? And why are they different from the results of your genome analysis?
40. Why are the results of your genome analysis different from the census?
41. What is the difference between a Vahaduo Single analysis and a Vahaduo Multi analysis?
42. What is the difference between proximal genome ancestry and distal genome ancestry?
43. Why do you prefer Global25 over other genetic analysis technologies?
44. I am a longtime reader of your blog. You have regularly made predictions that turned out to be true. What is your secret?
45. I like this website. How can I financially contribute?
46. How can I contact you?

What is this website about?

This website is dedicated to all things pertaining to the mysterious Land of Punt — perhaps the first resource of its kind. That includes the ancient territory and its inhabitants (the Puntites), as well as the history, culture, anthropology, genetics, linguistics and archaeology of Punt’s descendant populations.

Where was the ancient Land of Punt?

The totality of the evidence on Punt locates this ancient territory in Northeast Africa, in a broad area extending from Cape Guardafui in the northeastern Puntland region of Somalia to the Red Sea Hills in northeastern Sudan.

For more details, refer to Punt: an ancient civilization rediscovered.

Who were the Puntites?

The Puntites appear to have been early Cushitic speakers. This is suggested by a number of things; notably, the ethnonym brbrta, which the ancient Egyptians reserved for the Puntites. The Periplus of the Erythraean Sea, a guide written in the 1st century CE by an Alexandrian merchant, indicates that peoples of the same name (“Berbers”) inhabited the Horn and Nile Valley, in areas exactly coinciding with the geographical distribution of most modern Cushitic-speaking populations.

Why did the ancient Egyptians refer to the Puntites as brbrta if they were Cushitic speakers rather than Berber speakers?

The ancient Egyptians derived the appellative brbrta from the sound of the Puntites’ language, which is thought to have contained a lot of “ber” and “bar” sounds. Hence, in allusion to this tongue, the New Kingdom Egyptians referred to the Puntites as Berbers. This is the earliest documented usage of the term.

What evidence is there that the Puntites spoke an Afro-Asiatic language?

During the 18th Dynasty expedition to Punt, which the female Pharaoh Hatshepsut organized, the ancient Egyptians communicated with their Puntite congeners in a common language. This fact is clearly depicted at Hatshepsut’s monument in Deir el-Bahri, Egypt. Hieroglyphics on the temple walls show Parahu (Perahu), a chief of Punt, conversing with the Egyptian envoys about the purpose of their voyage to Punt. Ergo, the ancient Egyptians could understand and speak the Puntites’ native tongue, without requiring an interpreter. This is consistent with the suggestion that the Puntites spoke an Afro-Asiatic language closely related to that of the Egyptians (similar to how many Brazilians today can comprehend the Spanish spoken by their neighbors, and many Colombians, Argentinians, Venezuelans, etc. can likewise grasp the related Portuguese language of the adjacent Brazilians).

What is the etymology of the term “Punt”?

The term Punt is thought to have been derived from Opone (Opun), one of the old “Berber” ports described in the Periplus. This emporium was located at Ras Hafun in northeastern Somalia, an archaeological site from where Roman artifacts and other objects of trade have been excavated.

Could the ancient “Puntite” statues that were excavated in Somalia actually be forgeries?

That is, of course, theoretically possible. However, the lack of demand for “Puntite” antiquities makes this unlikely since there would be no monetary incentive to create elaborate forgeries. Moreover, very similar sculptures have been excavated further north in the Nile Valley, at Tombos. This makes the prospect of fabrication that much less probable.

See The Mystery of the Land of Punt Unravelled – book review for a fuller discussion on these artifacts.

Were the ancient Egyptians related to the Puntites?

Yes. According to the ancient Egyptians themselves, the Puntites were close relatives of theirs. The kings of Egypt’s 1st Dynasty also claimed to have originated from Punt, as did the female Pharaoh Hatshepsut, who reigned during the 18th Dynasty.

What is the genetic composition of the modern Afro-Asiatic-speaking populations of the Horn of Africa?

The modern Cushitic, Ethiosemitic and North Omotic-speaking populations of the Horn region have a predominant non-African ancestry (over 70%), which consists of majority West Eurasian elements (ancient Egyptian, European Steppe, and Levantine Natufian components) and a minority East Eurasian element (East Asian component). They also bear minor Sub-Saharan African admixture (~27%) and a residual North African Iberomaurusian/Taforalt admixture (under 3%). Muslim Egyptians and Libyans have a similar ancestral makeup, with additional recent gene flow from western Asia. Coptic Egyptians and northern Egyptians (from Cairo and Mansoura) descend directly from earlier Dynastic period Egyptians, with few influences from outside sources. Maghrebis have a comparable ancestral makeup as Libyans, but carry a greater Iberomaurusian ancestry (~29%); they also harbor notable Anatolian Neolithic admixture (~23% on average), particularly in the northern coastal areas opposite the Iberian peninsula.

See our genome analysis Genetic affinities of the Cushitic, Ethiosemitic and North Omotic-speaking populations of the Horn of Africa for details.

Do the ancient Cushites share this ancestral makeup?

Yes. The Cushites of the Pastoral Neolithic have an almost identical ancestral composition as modern Cushitic, Ethiosemitic and North Omotic-speaking individuals of the Horn region.

Do modern South Omotic speakers share this ancestral makeup?

No. The Ari and other South Omotic-speaking populations are physically, genetically and culturally distinct from their North Omotic, Ethiosemitic and Cushitic-speaking neighbors. South Omotic speakers are instead similar to Nilo-Saharan/Niger-Congo-speaking groups. This suggests that they originally spoke Nilo-Saharan/Niger-Congo languages, and later switched to speaking Afro-Asiatic languages.

See Ancient DNA from Ethiopia for more on this.

How come there aren’t any Northern Somali, Saho, Tuareg Berber or Nubian samples in your Vahaduo genome analysis?

Currently, there are no Northern Somali, Ethiopian Somali or Djiboutian Somali samples listed on the official Global25_PCA_modern dataset. All of the Somali coordinates contained therein belong to either Southern Somali or Kenyan Somali individuals. However, some genotype raw data files of Northern Somali individuals from the northeastern Puntland region of Somalia have now been made publicly available here on the BioStudies website. We are presently trying to convert that raw data into Global25 coordinates, so that we may add Northern Somali individuals to our Vahaduo Admixture JS genome analysis.

Likewise, there aren’t any Nubian samples listed on the official Global25_PCA_modern dataset. Hollfelder et al. (2017) published Nubian genotype data, which some users online have converted into Global25 coordinates. However, there apparently was an error in the conversion process since the Mahas and Danagla Nubian specimens display extremely high Distance fits and other signs typical of low quality sampling (this is explained in greater depth here).

As regards the Saho and Tuareg Berbers, they too are not listed on the official Global25_PCA_modern datasheet. There, in fact, doesn’t seem to be any published genotype data on these two populations. As soon as we locate such raw files, we will endeavor to convert them to Global25 coordinates for incorporation into our Vahaduo genome analysis.

If the Cushitic, Ethiosemitic and North Omotic-speaking populations of the Horn really do have a predominant non-African ancestry, why do some analyses claim otherwise?

The analyses which claim that the Cushitic, Ethiosemitic and North Omotic-speaking populations of the Horn have less than 70% non-African ancestry underestimate the non-African ancestry that these groups actually bear. These works are all characterized by one or both of two major shortcomings: 1) they limit their reference samples to two or three proxy groups, and/or 2) they do not include “pure” ancient reference samples among those proxies. Predictably, the end result is that these analyses fail to capture much of the non-African ancestry that the Horn’s Afro-Asiatic speakers do, in fact, carry.

The Cushitic, Ethiosemitic and North Omotic-speaking individuals actually require four ancient Eurasian samples (Egyptian, European Steppe, Levantine Natufian and East Asian specimens), two ancient Sub-Saharan African samples (Nilo-Saharan and East African hunter-gatherer specimens), and one ancient North African sample (Iberomaurusian specimen) to fully encapsulate their ancestral composition. We know this because when a Single analysis is conducted on these samples using the Vahaduo Admixture JS program, the aforementioned components are what actually show up as their ancestral makeup, and always at frequencies averaging at least 70% non-African ancestry.

This fact can be easily visualized in the Vahaduo Multi table below (Eurasian components: ancient Egyptian=EGY_1879BCE, European Steppe=RUS_Baikal, Levantine Natufian=Levant_Natufian_EpiP, ancient East Asian=CHN_Huatuyan_500BP ; Sub-Saharan African components: ancient Pygmy=CMR_Shum_Laka, ancient Nilo-Saharan=Kakapel_300BP, ancient East African Hunter-Gatherer=MWI_Chencherere ; North African component: Iberomaurusian=MAR_Taforalt):

Genome analysis on the Vahaduo Admixture JS program indicates that the contemporary Cushitic/Ethiosemitic/North Omotic speakers of the Horn carry a predominant non-African ancestry, which, on average, comprises over 70% of their ancestral makeup. This non-African ancestry can be further broken down into majority West Eurasian elements (ancient Egyptian, European-related Steppe, and Levantine Natufian components) and a minority East Eurasian element (ancient East Asian component). Additionally, these Afro-Asiatic-speaking individuals bear some low-to-moderate Sub-Saharan African admixture (~27%), as well as a very minor North African Iberomaurusian admixture (under 3%).

Genome analysis has also confirmed that the Cushites of the Early Pastoral Neolithic — who are the oldest Cushitic individuals so far to be genetically analysed — had just ~8.5% ancient Nilo-Saharan admixture. The rest of their Sub-Saharan African admixture consists of the ancient East African hunter-gatherer component (~14.7% on average). Because this forager element is indigenous to eastern Africa, this implies that the earliest Cushitic settlers would have almost entirely borne non-African ancestry at the time of their first arrival in the region (~91.5%):

Early Pastoral Neolithic (Kenya)

Can Cushitic, Ethiosemitic and North Omotic-speaking individuals be genetically modelled as a simple two-way admixture of West Eurasians and Sub-Saharan Africans?

Cushitic, Ethiosemitic and North Omotic speakers can be modelled as a two-way admixture between West Eurasian and Sub-Saharan African populations, but doing so is unrealistic. For example, in both the Vahaduo Admixture JS and qpAdm programs, these individuals can be constructed as a mixture between Levantine Natufian and ancient Nilo-Saharan populations. However, this genetic model fails to account for many West Eurasian-associated SNPs that the Cushites of the Pastoral Neolithic and modern Afro-Asiatic speakers from the Horn carry and which the Natufians did not and thus could not have transmitted (e.g. the derived alleles at the SLC24A5 and APBA2 genes, which code for lighter skin pigmentation).

A more realistic two-way admixture model would use as its proxy groups an ancient Egyptian sample (e.g. the official Global25 specimen Late_Period:JK2134) and a “pure” ancient Nilo-Saharan sample (Kakapel_300BP). As can be seen in the Vahaduo Multi table below, such a construction produces a similar overall ancestry apportionment as our genome analysis. The Afro-Asiatic speakers again carry a predominant Eurasian ancestry (averaging 77%), with a minority Sub-Saharan African admixture (23%):

The two-way admixture model above uses an ancient Egyptian source as a West Eurasian proxy instead of a Levantine Natufian source. As a result, it does a better job at outlining the ancestral composition of the Horn’s Afro-Asiatic speakers. This fact is also reflected in this model’s improved average Distance fit over that of the Natufian two-way admixture model. Nonetheless, the ancient Egyptian two-way admixture model is still unsatisfactory because it does not explain why some Cushitic speakers bear European Steppe-affiliated mtDNA clades (e.g. the I2 maternal haplogroup). This simplistic framework, moreover, does not capture East Eurasian-associated genetic variants that these Afro-Asiatic-speaking individuals are known to possess (e.g. the EDAR allele discussed below). It also does not take into consideration particular mutations that the Afro-Asiatic speakers acquired through contact with neighboring hunter-gatherers (e.g. variants associated with high altitude adaptation, which many Oromos, Wolayta and Abyssinians now carry).

Our genome analysis does all of the above, and is therefore the most reliable and comprehensive genetic model.

Can the ancestral composition of the Cushitic, Ethiosemitic and North Omotic-speaking populations realistically be modelled in qpAdm with a five-way admixture scenario?

For the most part, yes. However, you would need to use as your Eurasian proxies an ancient Egyptian source (the Late_Period:JK2134 or Levant_Beirut_IAIII_Egyptian:SFI-44 samples) and a source carrying both European Steppe ancestry and ancient East Asian ancestry (the TUR_Ottoman:MA2196 specimen from Central Asia, which is discussed further below). For your Sub-Saharan African proxies, you would require a “pure” ancient Nilo-Saharan source (Kakapel_300BP) and a “pure” ancient East African hunter-gatherer source (MWI_Chencherere).

Such a qpAdm genetic model can produce a realistic ancestral breakdown for Afro-Asiatic speakers from the Horn and for northern Sudanese. However, it is inadequate for Muslim Egyptians and Libyans, as these individuals require further proxies to capture their extra Natufian, Anatolian Neolithic and Iran Neolithic admixtures. Libyans also need an additional reference population to accurately quantify their greater North African Iberomaurusian admixture (MAR_Taforalt).

If the Cushitic, Ethiosemitic and North Omotic-speaking populations of the Horn really do have a predominant non-African ancestry, why is this not more clearly reflected on a PCA map?

In terms of genetics, the position of a sample on a Principal Component Analysis (PCA) plot is dictated by two principal factors: 1) the frequency or percentage of a genetic component(s) which that sample carries, and 2) the relative degree of divergence of said component as measured by genetic distance.

For example, an individual with an Irish parent and a Yoruba parent will genetically plot differently from another individual with a Sardinian parent and a Yoruba parent, despite the fact that both persons have a European parent and a Yoruba parent. Although both individuals will sit in an intermediate position between unmixed European individuals and unmixed Yoruba individuals, the Sardinian-Yoruba individual will naturally be positioned closer to unmixed Yoruba individuals compared to the Irish-Yoruba individual. This is not because the Sardinian-Yoruba person is “more” Sub-Saharan African or “less” European than the Irish-Yoruba person. Rather, it boils down to differences in the relative proportions of the Irish’s and Sardinians’ three signature European ancestral components (Steppe, Anatolian Neolithic and Western Hunter-Gatherer elements), which each have different genetic distances from the Yoruba’s signature Sub-Saharan African component (COG_Kindoki element).

Genome analysis using the Vahaduo Admixture JS program has established that the minority Sub-Saharan African admixture, which the Afro-Asiatic speakers in the Horn bear, primarily consists of an ancient Nilo-Saharan element (best represented by the Kakapel_300BP sample) and an ancient East African Hunter-Gatherer element (best represented by the MWI_Chencherere sample). The Chencherere forager cohort from Malawi is the second most divergent of all existing genome components. Only ancient Khoisan samples from southern Africa (best represented by the ZAF_2000BP specimens) have a greater genetic distance from modern individuals. Thus, while the Cushitic, Ethiosemitic and North Omotic-speaking populations do bear over 70% non-African ancestry, the fact that the highly divergent East African Hunter-Gatherer element constitutes a notable portion of their remaining ~27% Sub-Saharan African admixture has an inordinately deviating effect on their position on a PCA map.

Moreover, whether or not an individual or population bears significant East Eurasian admixture also affects their location on a PCA plot. East Asians and many Central Asians, for example, are typically concentrated on a different PCA axis from Europeans. This is because their patented East Eurasian ancestral components have a greater genetic distance from the trademark European genetic components. Therefore, groups or persons that carry such East Eurasian ancestry — which we now know includes the Cushitic, Ethiosemitic and North Omotic-speaking populations of the Horn — will have a pull toward that axis. This further complicates the positioning of said Afro-Asiatic speakers on a PCA map, precluding simplistic interpretations.

Were the ancient Azanians Cushites?

Genetic analysis, cultural evidence and antiquities recovered from Azanian sites suggest that indeed they were.

For specifics, refer to Who were the ancient Azanians?.

If Muslim Egyptians have other admixture elements besides the ancient Egyptian component, does this mean they are mostly of non-Egyptian origin?

Ancient DNA data thus far accumulated indicates that there were two basic communities in earlier Dynastic period Egypt: a northern group centered in Lower Egypt comprising the original Egyptian population (as represented by the Middle Kingdom nobleman Nakht-Ankh, exhumed at Deir Rifeh in Lower Egypt), and a southern group centered in Upper Egypt consisting of newcomers from Nubia (as represented by the Amarna royal family, who reigned from Amarna in Upper Egypt). The Lower Egyptian group shared affinities with the Pre-Pottery Neolithic makers of the Levant, while the Upper Egyptian group shared affinities with European Steppe and East Asian populations.

With the progress of time, these two originally distinct populations would intermingle, thus forming the core of the later Dynastic period Egyptian profile (as represented by the Late Period ancient Egyptians from Abusir, a site in Middle Egypt). Muslim Egyptians descend from these later Dynastic period Egyptians (with some extra recent admixture from western Asia), whereas Coptic Egyptians and northern Egyptians descend from the earlier Dynastic period Egyptians.

You indicate that you used the Nakht-Ankh cohort to arrive at your conclusion that Afro-Asiatic speakers from the Horn and North Africa carry some European Steppe and East Asian admixture. Since Nakht-Ankh is not an official Global25 sample, could it be that that is the actual reason behind this result?

When conducting a Single analysis on the Vahaduo Admixture JS program, the modern Afro-Asiatic-speaking populations of the Horn of Africa, Nile Valley and (to a lesser extent) Maghreb do always bear some European Steppe and East Asian admixture. This remains true whether or not we add Nakht-Ankh’s unofficial EGY_1869BCE sample to the official Global25_PCA datasheet. Multiple other lines of evidence also corroborate this finding, including the autosomal STR markers of the Amarna royal family.

Is it true that the ancient Egyptians, like the Minoans, descend from Anatolian Neolithic farmers?

There indeed is an Anatolian Neolithic element in the ancestral composition of ancient Egyptians. However, it is embedded within the Egyptians’ earlier Dynastic period component, as represented by the specimen Nakht-Ankh. This element consists of mostly Levantine Natufian ancestry (~80%), with a minor Anatolian Neolithic admixture (~20%). It is thus similar to the ancestral makeup of the Pre-Pottery Neolithic culture bearers of the Levant, albeit with considerably less Anatolian Neolithic admixture (PPN makers have around 50% Natufian and 50% Anatolian Neolithic ancestries).

Due to this difference in admixture ratios, we cannot, given our current state of knowledge, say for certain whether the Anatolian Neolithic element arrived in the Nile Valley from the Levant, the Aegean or elsewhere. It is possible that it entered the Egypt area directly from southern Europe. This is because Anatolian Neolithic admixture has been detected in Late Neolithic specimens excavated at the Kelif el-Boroud site in Morocco, where the component is associated with introgression from the Cardial Mediterranean Neolithic culture of Iberia.

Did the Capsians of North Africa descend from the Iberomaurusians?

At present, we do not have any ancient DNA from Capsian sites. However, craniometric analysis of Capsian and Iberomaurusian specimens does suggest that they were closely related. The Capsians therefore appear to be a sibling population to the Iberomaurusians’ immediate descendants, the Early Neolithic peoples excavated at the Ifri n’Amr or Moussa site in Morocco.

This association is also supported by preliminary reports from an upcoming paleogenetic study, which includes new samples of Mesolithic and Neolithic specimens exhumed in Tunisia (see slide here). These ancient individuals cluster in a very similar way vis-a-vis modern Maghrebi populations as do the Mesolithic Iberomaurusians and Early Neolithic Moroccans, suggesting that the Neolithic Tunisians may represent the first veritable Capsian specimens to be genomically analysed.

What are the genetic affinities of the Aterians?

As of 2023, the prehistoric Aterians of North Africa (c. 40,000BP) have not been genetically analysed. However, craniometrically, they have been found to be distinct from the Iberomaurusians and Capsians.

It has been hypothesized that the Aterians may have had a Khoisan-related affiliation. This assumption is not supported by DNA analyses of Epipaleolithic Iberomaurusian specimens, Early and Late Neolithic individuals in Morocco, and contemporary Maghrebi populations, which indicate no ties with ancient or modern Khoisan samples. The minor Sub-Saharan African admixture that is carried by Afro-Asiatic speakers in Northwest Africa seems instead to be primarily associated with Niger-Congo speakers and secondarily with Nilo-Saharan speakers. Hence, the Aterians may have been ancestral to modern Niger-Congo-speaking groups inhabiting West Africa.

Are you sure that the oblique eye folds, which are common among Afro-Asiatic speakers and Nubians in the Horn and North Africa, are not due to absorption of Khoisan peoples? What makes you ascribe this trait to Central Asian or East Asian admixture?

Scientists have observed a positive correlation among Khoisan individuals in the occurrence of their distinctive “narrow eye opening” and “peppercorn” hair. In this population, these traits appear to be controlled by the rs111298318 allele of the KRT74 gene. Fan et al. (2023) report that this rs111298318-C mutation is carried by most Khoisan samples they analysed (>73%), whereas it is rare among other populations of Africa and Eurasia (<5%).

Furthermore, among Afro-Asiatic speakers from the Horn region, there is a disconnect between the incidence of “narrow eye opening” and “peppercorn” hair. Gallo (1979) notes that, although most adult male and female individuals in his southern Somalia dataset had normal eye folds (36.6%), a high proportion of this cohort had either oblique eye folds (29.9%) or slight Mongolian eye folds (19.9%) and the rest had other eye folds (13.6%). However, in contrast to the Khoisan in Southern Africa, there is a near zero percent incidence of “peppercorn” hair among the Cushitic-speaking populations of Northeast Africa (cf. Charpin and Georget (1977); Coon (1939)). Thus, the allele(s) responsible for the oblique and Mongolian eye folds among these Afro-Asiatic speakers is different from the rs111298318 mutation which apparently causes expression of both the “narrow eye opening” and “peppercorn” hair among Khoisan individuals.

Additionally, genome analysis of contemporary Afro-Asiatic-speaking populations in the Horn, Nile Valley and Maghreb has found almost no trace of a Khoisan genome influence. On the other hand, all of these groups, except Coptic Egyptians, bear varying degrees of an East Asian-related ancestral element (best represented by the CHN_Huatuyan_500BP sample).

What’s more, there is other genetic evidence of contact with East Asian populations, either directly or indirectly. Notably, a significant minority of modern Cushitic individuals carry the derived allele of the EDAR gene, which is associated with hair thickness among groups in eastern Asia (up to 12.5% of Somalis bear this mutation; cf. ALFRED).

Is there any other evidence of an ancient Central Asian or East Asian-related presence in the Nile Valley?

Yes.

Among Afro-Asiatic speakers:

Ancient Egyptian monarchs belonging to the Amarna royal family, which reigned from Upper Egypt during the 18th dynasty, were analysed for autosomal STR markers. These rulers were found to share greatest affinities with modern individuals from South Asia, Europe (particularly eastern Europe), and Northeast Asia (cf. LOP (2023)). Furthermore, haplogroup analysis has detected the Y-DNA clade F-M89 among Christian-era specimens excavated at Meroe Island and along the 4th Cataract in Sudan (Yousif and Eltayeb (2009)). This is the earliest reported occurrence of this lineage in Africa. The F-M89 paternal clade is today mainly restricted to populations in South/Central Asia, which points to an ancient migration(s) from this area into Northeast Africa.

Among Nilo-Saharan speakers:

A Vahaduo Distance analysis performed on the Nilo-Saharan-speaking samples listed on the official Global25_PCA_modern datasheet reveals that all but one of these individuals have the TUR_Ottoman:MA2196 sample as their top Eurasian result (see here). This means that, out of the thousands of ancient Eurasian populations listed on the Global_25_PCA datasheet, that is the specimen they share the most immediate affinity with (as measured in terms of genetic distance). It is not with the neighboring ancient Egyptians or peninsular Arabs, as one might have expected. Moreover, in line with these Central Asian ties, the aforementioned haplogroup F-M89 has been detected at elevated frequencies among the Berta (55.55%) and Gumuz (33.33%), Nilotic peoples who inhabit the border area between Ethiopia and South Sudan (cf. Balemi (2018)). This suggests that, like their Afro-Asiatic-speaking neighbors, these Nilotic groups would also have come into contact with and absorbed some newcomers from Central Asia.

If we conduct a Vahaduo Distance analysis to find out what are the TUR_Ottoman:MA2196 sample’s closest modern associations, we see that the specimen shares nearest affinity with Turkmen, Uzbeks and other similar populations inhabiting Central Asia (see here). These individuals are characterized by a mixture of European Steppe and East Asian-related ancestry. This confirms the results of our genome analysis, which has identified an ancient presence in the Nile Valley of peoples from South/Central Asia.

What complexion were the ancient Cushitic settlers of eastern Africa?

The information we have about the skin pigmentation of the early Cushitic settlers in eastern Africa is gleaned from three main sources: 1) archaeogenetic analysis of ancient Cushitic individuals, 2) genetic analysis of modern Cushitic-speaking individuals, and 3) ancient iconography.

Wang et al. (2020) examined two of the ancient Cushitic settlers of the Pastoral Neolithic for phenotype alleles; a 2300 year old sample from Hyrax Hill and a 1500 year old sample from Molo Cave, archaeological sites located in Kenya. The scientists report that both of the specimens bore the derived Ala111Thr (rs1426654) allele of the SLC24A5 gene, which confers a lighter complexion (cf. Table S7). Of the known genetic variants controlling skin pigmentation (which number over 20), this derived SLC24A5 mutation has been measured to have the single greatest effect on coloration: 7.6-11.4 average melanin units for the derived SLC24A5 allele, compared to just 5 average melanin units for the derived SLC45A2 allele and 2.8-3.8 average melanin units for the derived KITLG allele (Beleza et al. (2013)). Besides the derived SLC24A5 variant, Wang et al. observed that their ancient Cushitic samples carried the derived allele at the APBA2 locus, which is likewise associated with lighter skin pigmentation. Altogether, ancient DNA analysis suggests that the early Cushites would have been of a relatively fair complexion, though just how light-skinned is uncertain since these individuals were not analysed for the other, less impactful pigmentation-related alleles. This, in turn, means that the ancient Cushites did not inherit their lighter skin-conferring mutations from the Levantine Natufians because the Natufians did not have any such alleles. 

Genetic analysis of modern Cushitic, Ethiosemitic and North Omotic speakers of the Horn indicates that most carry the derived SLC24A5 allele (~60%; cf. ALFRED; Tekola-Ayele et al. (2015)). These Afro-Asiatic-speaking individuals have not been studied for the derived APBA2 mutation of the early Cushites. However, they have been found to bear low frequencies of the derived SLC45A2 allele (0%-1.4% of Ethiopian Jews; 0%-3.3% of Somalis), as do peninsular Arabs (10.5% of Qataris; 11.6% of Saudis) (cf. ALFRED). This is similar to ancient Egyptian specimens excavated at the Abusir site in Middle Egypt, two of whom Schuenemann et al. (2017) assert carried the derived allele for the SLC24A5 gene but not the derived SLC45A2 mutation. The derived SLC45A2 allele seems instead to have been mainly spread in the Levant and Arabian peninsula during the later Bronze Age by outsiders from the Caucasus/Iranian Plateau since a) Bronze Age specimens exhumed at Sidon, Lebanon, who are otherwise very similar to modern Lebanese individuals, did not bear this variant (Haber et al. (2017)), and b) this mutation is today found at elevated frequencies among Levantine populations that have appreciable Caucasus Hunter-Gatherer/Iran Neolithic admixture (Druze, Samaritans).

Lastly, the artistic representations which the ancient Egyptians left of their Puntite relatives affirm that the latter were of a copperish hue. Puntite figures are most notably depicted at Pharaoh Hatshepsut’s monument at Deir el-Bahri and on the Grand Procession mural in the tomb of Rekhmire at Thebes, and always in a manner akin to the Egyptians themselves.

Were the ancient Cushites slender?

Yes. Ancient artwork informs us that the early Cushites had a linear physique. A mural has been discovered in Egypt, which dates to the 12th Dynasty and depicts an ancient Cushitic man with a slim build, fine features, an orthognathous profile and loose wavy hair.

Additionally, examination of the bones of ancient Cushitic settlers in eastern Africa has established that, on average, they had a slender constitution. This was a distinctive characteristic they shared with their close relatives the ancient Egyptians. In his capacity as chief anatomist for the Archaeological Survey of Nubia, the Egyptologist Grafton Elliot Smith inspected thousands of predynastic Egyptian specimens. He also studied many predynastic Egyptian skeletons excavated from the Naqada site by the Hearst Expedition under George Andrew Reisner, as well as predynastic Egyptian remains recovered by Randall-MacIver at El-Amrah near Abydos. According to Elliot Smith (1911):

The hot, dry sands of Egypt have preserved through a span of more than sixty centuries the remains of countless multitudes of the earliest people known to have dwelt in the Nile Valley ; and not the mere bones only, but also the skin and hair, the muscles and organs of the body ; and even such delicate tissues as the nerves and brain, and, most marvellous of all, the lens of the eye, are available for examination to-day. Thus we are able to form a very precise idea of the structure of the body of the Proto-Egyptian.[…]

Our information concerning these earliest inhabitants of the Nile Valley has been acquired from the study of the contents of many thousands of their graves, found in cemeteries scattered in every part of Egypt and Nubia so far examined.[…]

[The Proto-Egyptian] was of very slender build, for his bones are singularly slight and free from pronounced roughness and projecting bosses that indicate great muscular development. In fact, there is a suggestion of effeminate grace and frailty about his bones, which is lacking in the more rugged outlines of the skeletons of his more virile successors.

Thus, modern Cushitic, Ethiosemitic and North Omotic-speaking individuals of Northeast Africa simply inherited their linearity from their Cushitic ancestors. We know with some certainty that this physique was not acquired from ancient Nilo-Saharan peoples. For starters, anthropometric analyses have shown that contemporary Afro-Asiatic speakers from the Horn (viz. Somalis, Oromos/Gallas, Afars, Amhara) cluster with Egyptians, Nubians and Yemenis rather than with Nilo-Saharan speakers from the Nile Valley (Mabaan, Nuba, Dinka, Shilluk, Anuak) (cf. (Billy (1988); Leguebe (1981)).

Furthermore, Melton et al. (2002) found that Cushitic Somalis have lumbar spine and femoral neck bone mineral content, bone mineral density, and bone mineral apparent density that is overall more similar to that of European Americans than to that of African Americans. By contrast, individuals from south Sudan (most of whom are Nilotes) have lumbar spine bone mineral density that is significantly lower than that of both European Americans and African Americans (Gong et al. (2006)).

Anthropometric analysis of Afro-Asiatic and Nilo-Saharan-speaking populations. The Cushitic-speaking samples (Galla (G) from central Ethiopia and Oromo (O) from west-central Ethiopia) group with Egyptians from the Kharga Oasis (K1, K2). The Nilo-Saharan-speaking samples form their own seperate cluster. This agrees with the view that the Land of Punt was situated in the Horn region (Leguebe (1981)).

Anthropometric relations between ethnolinguistic groups in Africa and the Arabian peninsula. The Cushitic and Ethiosemitic-speaking populations in the Horn of Africa cluster with other Afro-Asiatic speakers in North Africa (northern Libyans and Egyptians) and the Arabian peninsula (Yemenis). Nubians also group with the Afro-Asiatic-speaking samples rather than with the Nilo-Saharan and Niger-Congo-speaking populations. This again supports a Red Sea area location for ancient Punt (Billy (1988)).

Lumbar spine and femoral neck bone mineral content, bone mineral density, and bone mineral apparent density among Somali, White and African American individuals. Overall, the Somali samples have osteological values that are more similar to those of the White samples than to those of the African American samples (Melton et al. (2002)).

For additional details, see The Elongated African fallacy.

Did the ancient Cushites have soft-textured hair?

Yes. Forensic examination of the ancient peoples of the Kerma civilization in Sudan indicates that, except for a few assimilated individuals of Nilotic origin, they uniformly had non-kinky hair texture. Lepisus (1915) notes that “most of the men, especially the principal skeletons, had fine heads with straight black hair.”

Furthermore, Wang et al. (2022) genetically analysed one such Kerma period specimen, who was excavated at the Kadruka site in Upper Nubia (northern Sudan). This individual had soft-textured wavy hair. His genome was found to be “genetically indistinguishable from that of early Neolithic eastern African pastoralists located 2500 kms away.” Ergo, the Cushites of the Pastoral Neolithic, who bore a predominant Eurasian ancestry (see above), simply passed on their soft-textured hair to their modern descendants.

Ancient Cushitic man, depicted on a 12th Dynasty monument. Note the figure’s sharp features, slim build, and loose, wavy hair. An individual excavated at the Kadruka site in Upper Nubia, which dates from the same time period (Kerma epoch), was found to have similar hair form. This specimen’s genome was reported to be “genetically indistinguishable from that of early Neolithic eastern African pastoralists located 2500 kms away,” confirming that the Cushites of the Pastoral Neolithic had soft-textured hair.

A Cushitic Bisharin Beja man with soft-textured wavy hair. Forensic, genetic and iconographic analysis indicates that the Cushites of the Pastoral Neolithic had non-kinky hair, which they passed on to their modern descendants.

A soft-textured wavy hair lock belonging to the Kerma-period ancient Cushitic individual, who was exhumed from the Kadruka site in Upper Nubia (northern Sudan) (Wang et al. (2022)).

Have the Kiffians and Tenerians of the Gobero been genetically analysed yet?

No, they have not. The Sereno lab, which carried out the excavations that identified these remains, indicates that the Kiffians and Tenerians were being sequenced for ancient DNA. However, that was a while ago (circa 2008), and no paleogenetic study has been forthcoming. There also has been little news as to when such an analysis will be conducted. Hopefully, the project has not been shelved.

Is it true that the ancient Egyptian Pharaoh Ramesses III and the Prince Pentawere (Unknown Man E) belong to the E1b1a haplogroup?

No. These royals actually belong to the E1b1b haplogroup. The misconception that Ramesses III and Unknown Man E bear the E1b1a clade is due to an incomplete analysis of their Y-STR markers.

Hawass et al. (2012) originally typed these monarchs for a number of Y-STR markers, including GATAH4. However, Whit Athey’s Haplogroup Predictor, the tool which Hawass et al. indicate they used to assign those Y-STRs to a specific paternal haplogroup, does not have an option for GATAH4. Consequently, Hawass et al. could not input the royals’ GATAH4=13 marker into that haplogroup predictor. This incomplete set of Y-STRs — that is, all of Ramesses III’s and Unknown Man E’s markers, minus GATAH4 — is what actually produces the false E1b1a result.

When the complete set of analysed Y-STR markers are inputed into the separate Nevgen Haplogroup Predictor (a tool which, in contrast, does have an option for GATAH4), these royals are instead assigned to the V22 subclade of the E1b1b lineage. This is a common haplogroup today in Egypt. E1b1b-V22 has also been observed among ancient Egyptian mummies at the Kurchatov Institute, further confirming that this is the correct haplogroup assignment. 

Is it true that the ancient Egyptian Amarna royal family shares closest genetic affinities with Sub-Saharan African populations?

No. This misconception is due to a small autosomal STR database, against which the Amarna royals’ microsatellite markers were compared.

Hawass et al. (2010) typed these 18th Dynasty Egyptian monarchs for autosomal short tandem repeats (autosomal STRs). The genetic testing company DNA Tribes then compared these specimens’ microsatellite markers with those belonging to various modern populations contained within its internal database, and reported that they showed greatest genetic affinity with those of contemporary Sub-Saharan African individuals. However, a cross-analysis of the Amarna royals’ autosomal STRs with the microsatellites listed on the more extensive Allele Frequency Database (ALFRED) demonstrates instead a close affiliation with populations in South Asia and Europe. This is consistent with the European Steppe presence discussed above, which we now know existed in the ancient Nile Valley.

For details, see our study Autosomal STR Analysis of the Ancient Egyptian Amarna Royal Family, Pharaoh Ramesses III, and Unknown Man E (Prince Pentawere).

In your autosomal DNA analysis of the ancient Egyptian Amarna royal family, could the South Asian affinity you detected actually be related to the Andamanese Hunter-Gatherers (Ancestral South Indians) instead of European Steppe populations?

That is unlikely. For starters, we observed a positive correlation between South Asian affinity and European affinity: the alleles with a primary South Asian affiliation also often included among their top results the Allele Frequency Database’s eastern European samples, and eastern Europe is where the Steppe component is believed to have ultimately originated (e.g. 22.50% of Croatians carry the Pharaoh Akhenaten’s FGA=23 allele, which peaks at 40% in a Reddy/Vanne sample from South Asia). We found the reverse also to be true in that the alleles with a primary European affiliation likewise frequently included South Asian groups among their top results (e.g. 48% of the Drokpa in South Asia carry the courtier Thuya’s D7S820=10 allele, which peaks at 52.60% in a Croatian sample).

Furthermore, while it is true that among contemporary South Asian populations in the ALFRED database the Reddy/Vanne show the most affinity with the Amarna royals, and modern Reddy/Vanne individuals do possess significant Andamanese Hunter-Gatherer ancestry, Kumar et al. (2022) report that they also bear some Steppe admixture (averaging ~10%). Since this is the only ancestral element that the Reddy/Vanne have in common with Croatians, the Steppe component appears to be what is causing the positive correlation. This finding is also concordant with the blondism found among the Amarna monarchs, as well as the Steppe-associated uniparental markers that they carry (viz. R1b-M296 paternal clade and I maternal haplogroup).

Moreover, in our Vahaduo Admixture JS genome analysis, we did not detect an Andamanese/Onge-related affinity in any of the ancient or modern Afro-Asiatic-speaking samples, including the ancient Egyptian specimens from Deir Rifeh and Abusir. The only exceptions are the medieval period Kulubnarti samples from Sudan. Several of these individuals do bear a trace Ancestral South Indian element of around 4%, in the form of the LAO_Hoabinhian component (the persons labeled I6327, I6325, I6255, I6251, I6250 and I18538 in the Vahaduo Single analysis here). However, because the Amarna era predates the Middle Ages by many centuries, we are concerned with an altogether different and much older population movement(s) from South/Central Asia into Northeast Africa — a migration(s) which almost certainly did introduce Steppe-related ancestry.

Which ancient DNA studies do you think should be published?

We welcome all ancient DNA studies; the more, the better. That said, some archaeogenetic works are more urgent than others. We would like to see papers published on the following early cultures, and preferably sooner rather than later:

Predynastic Egyptians (Tasian, Merimdian, Badarian, Naqadan), earlier Dynastic period Egyptians (Early Dynastic and Old Kingdom), later Dynastic period Egyptians, Mesolithic Nubians (Wadi Halfa, Khartoum), post-Neolithic Nubians (A-Group, C-Group, Kerma, Meroitic, Post-Meroitic/X-Group, Christian), ancient Libyans, Jebel Moya specimens, Jebel Sahaba specimens, Kiffians, Tenerians, Iron Age “Berbers” of the northern Horn, Iron Age “Azanians” of the southern Horn and Swahili coast, Garamantians of the Sahara, Capsians of North Africa, Aterians, Iwo Eleru specimen, Great Zimbabwe specimens, Himyarites, Sabaeans, other ancient specimens of the Arabian peninsula, Hyksos, Hittites, Sumerians, and Elamites.

It would be great to access the raw data from these would-be studies and get it converted to Global25 coordinates for detailed analysis.

Who were the ancient Nubians?

There were two distinct groups inhabiting ancient Nubia: the “red Nubians” and the “black Nubians”. These peoples were of different ancestral origins, much like how European Australians and Aboriginal Australians (though both Australians) are ultimately of different ancestral origins.

“Red Nubians” were the Egyptian-related inhabitants of ancient Nubia. They include populations like the Meroites and Post-Meroitic/X-Group makers. Their modern descendants include most Nubians, who live in the area between central Sudan and southern Egypt. Nubians are physically, genetically and culturally similar to Afro-Asiatic-speaking populations.

“Black Nubians” were the Nilotic-related inhabitants of ancient Nubia. Their modern descendants include groups like the Nuba (also spelled Noba), who live in the Nuba Mountains/Kordofan Mountains/Nuba Hills further south. Nuba are physically, genetically and culturally similar to Niger-Congo/Nilo-Saharan populations.

If Nubians are ancestrally related to Egyptians, why do they now speak a Nilo-Saharan language instead of an Afro-Asiatic language?

Nubians appear to have adopted their current Nilo-Saharan language (called Nobiin) some time around the Iron Age. This is the period marking the rise of the X-Group/Post-Meroitic culture, which succeeded the Meroitic civilization. Archaeogenetic data indicates that the X-Group peoples were of the same ancestral stock as the preceding Meroites, with a slightly greater Sub-Saharan African admixture. At present, the Meroitic language remains undeciphered. Linguistic evidence, however, suggests that it may have been an Afro-Asiatic language.

Refer to Ancient DNA from Sudan for more.

Are Sudanese “Arabs” really peninsular Arabs?

Generally-speaking, Sudanese “Arabs” are of the same ancestral origin as Nubians. They share close genetic, physical and (to some extent) cultural ties. Textual evidence also indicates that they spoke a common language as recently as the Middle Ages, during the spread of Islam. This is when Kababish, Gaalien, Arakien, Meseria and other Sudanese “Arab” groups are believed to have adopted their current Arabic language and Arabian genealogies.

Overall, the northern Sudanese (Nubians and Sudanese “Arabs”) have a similar ancestral composition as the Afro-Asiatic-speaking populations from the Horn. The Kababish “Arabs” of Sudan carry a predominant Eurasian ancestry (averaging almost 70%), which consists of majority West Eurasian elements (ancient Egyptian, European Steppe, and Levantine Natufian components) and a secondary East Eurasian element (ancient East Asian component). They also bear some Sub-Saharan African admixture (nearly 30% on average), which varies considerably per individual, much more than among the Afro-Asiatic speakers of the Horn. Certain Kababish individuals have a low SSA admixture of ~20% (e.g. the persons labeled KAB18, KAB36, KAB45, KAB46 on the chart below), whereas others have a high SSA admixture of ~50% (e.g. KAB12, KAB20, KAB22, KAB17).

Are Maghrebi “Arabs” really peninsular Arabs?

Most Arabic speakers in the Maghreb are Arabized Berbers. This has been confirmed genetically, as they carry the same overall signature Maghrebi ancestral components as Berber speakers. Having said that, there are some groups and individuals in Northwest Africa that clearly are descended from recent settlers from the Arabian peninsula. These communities include the Rbaya of Tunisia. Many individuals inhabiting the Douz area, also in Tunisia, likewise seem to be of recent peninsular Arab origin (Khaliji).

The recent peninsular Arab ancestry found among Northwest Africans (best represented by the Levant_Tell_Qarassa_Early_Antiquity component) can be seen in the table below. Among Maghrebi individuals, this element comprises an average of 11.7% of their ancestral makeup.

Are Ashkenazi Jews and Sephardic Jews really of the same ancestral origin as Mizrahi Jews, Palestinians, Lebanese, peninsular Arabs, Assyrians, Mahra and other Semites?

Yes. Ashkenazi Jews and Sephardic Jews descend from Middle Eastern Jews who settled in southern Europe. There, their Levantine ancestors intermixed with autochthonous Europeans and subsequently dispersed to other parts of the continent. This is why Ashkenazi and Sephardic Jews today have mostly Levantine-related Y-DNA or paternal DNA (derived from their Semitic male ancestors) and mostly southern European-related mtDNA or maternal DNA (derived from absorbed indigenous European women). In terms of genome affinities, this is also why modern Ashkenazi and Sephardic Jewish individuals on average bear around 50% Lebanese-related ancestry and 50% north Italian-related ancestry, like contemporary Maltese persons. We know from ancient DNA analysis of specimens buried at a Jewish cemetery in Erfurt, Germany, that this admixture process was still in its early stages as recently as the Middle Ages.

Furthermore, the Levantine roots of Ashkenazi Jews and Sephardic Jews are reflected in the languages they speak (Yiddish and Ladino, respectively). These idioms are creoles of the original Hebrew language spoken by their Levantine ancestors (an Afro-Asiatic tongue of the Semitic branch) and the native Indo-European languages of their European host communities.

STRUCTURE and phylogenetic analysis of Jewish populations. Ashkenazi Jews cluster with Mizrahi Jews from Syria and other Jewish groups. These Jewish populations carry both Levantine/Arabian (pink component) and European admixture elements (blue component). Ashkenazi individuals trace just over half of their ancestry to the Levantine/Arabian component (~60%), intermediate between Palestinians and southern European populations. This is consistent with the observation that the Semitic male ancestors of Ashkenazi and Sephardic Jews interbred with native women in southern Europe (Atzmon et al. (2010)).

Which commercial genetic test, if any, do you recommend?

We recommend that customers take a whole genome sequencing (WGS) test, as offered by companies like Dante Labs. However, since such tests are at the moment quite pricey, an acceptable alternative is the basic ancestry test offered by either of the two largest commercial testing labs, AncestryDNA or 23andme. These services are reasonably priced and will give a customer all the essentials, including that person’s Y-DNA haplogroup (for males only), mtDNA haplogroup, and genotype raw data.

Once a customer downloads his/her genotype file, we suggest that he/she takes advantage of Eurogenes’ offer to convert that raw data to Global25 coordinates (it’s cheap; costs only around $12). Through G25-compatible software like Vahaduo or nMonte, a user can then compare his/her genome with those of thousands of modern and ancient individuals listed on the official Global25 datasheets. This will allow the person to determine his/her ancestral composition with much greater accuracy than ever before.

For Afro-Asiatic speakers from the Horn of Africa and North Africa, utilize the coordinates below as your “pure” ancient Source populations in the Vahaduo Admixture JS program:

EGY_1879BCE:Nakht-Ankh,0.0012,0.129,-0.044,-0.0965,-0.0031,-0.0534,-0.017,-0.0078,0.0551,-0.0049,0.0138,-0.0172,0.0306,-0.0015,0.0069,-0.0072,-0.0111,0.0053,-0.004,0.0042,-0.0012,0.0046,-0.0078,0.0026,-0.0013 RUS_Baikal_BA:GLZ003,0.0033,-0.0361,0.0257,0.0068,-0.0349,-0.0148,-0.0004,0.0025,0.0041,-0.0037,0.0027,0.0003,0.0006,-0.0174,-0.004,-0.0063,-0.0059,-0.0029,0.0023,0.0131,-0.0249,0.0103,-0.0027,-0.0064,-0.0021 CHN_Huatuyan_500BP:HuatuyanNL04,0.0002,-0.0435,-0.0111,-0.0253,0.044,0.0103,-0.0045,0.0176,-0.0019,0.0001,-0.0068,0.0031,0.0004,0.0009,0.016,-0.0246,0.0165,-0.0256,-0.0042,-0.0184,0.0094,0.0232,0.0106,0.0186,0.0135 Levant_Natufian_EpiP:I0861,0.0017,0.0133,-0.0104,-0.0421,0.0087,-0.0273,-0.0081,-0.0107,0.0492,-0.0044,0.0176,-0.0131,0.0453,0.0012,0.0168,0.0197,-0.0079,0.0053,-0.0144,0.0331,-0.0033,-0.0026,-0.0116,-0.0093,0.01 MAR_Taforalt:TAF009,-0.0143,0.0076,-0.0065,-0.0245,0.0032,-0.0172,-0.0225,0.0089,0.065,0.0011,0.0156,-0.0178,0.039,-0.0385,0.0537,-0.0257,0.0018,-0.0444,-0.1038,0.0332,-0.021,-0.0973,0.0551,-0.0137,0.0103 KEN_Kakapel_300BP:KPL002,-0.0517,0.0057,0.0032,0.0026,-0.0003,0.0024,-0.0077,0.0079,0.0022,-0.0124,-0.0092,0.01,-0.0188,-0.0018,-0.0259,0.0185,-0.0089,-0.0011,-0.0115,0.0025,0.0007,-0.0016,0.0057,0.007,-0.0034 CMR_Shum_Laka:I10874_new_all,-0.0539,0.0055,0.0023,0.0094,-0.0001,-0.0011,0.0653,-0.049,0.0023,0.0033,0.0009,-0.0253,-0.0176,0.0012,0.0048,-0.0059,0.0063,0.0109,-0.0088,-0.0033,-0.0024,-0.0052,-0.0022,0.0017,-0.0039 COG_Kindoki_230BP:KIN002,-0.0543,0.0062,0.0054,0.0061,0.002,0.0071,0,0.0087,-0.0205,0.0094,0.0038,-0.0052,-0.0024,-0.0054,0.0028,0.0018,-0.0056,-0.0032,-0.0032,0.0089,0.0088,-0.0049,-0.0019,-0.0047,0.0069 MWI_Chencherere:I4421_new_all,-0.05,0.0072,0.002,0.0077,-0.0007,-0.0031,0.0672,-0.0583,0.0166,-0.0113,-0.0037,-0.0216,-0.0169,-0.0045,0.0111,-0.0107,0.0153,0.1034,-0.038,-0.0026,-0.0151,-0.0126,-0.0047,-0.007,0.0019 ZAF_2000BP:bab001,-0.0566,0.0058,0.0066,0.0108,-0.0004,-0.0015,0.1119,-0.0879,0.0074,0.0082,0.0064,-0.0451,-0.0141,0.0031,0.0186,-0.0136,0.0244,0.2362,-0.0785,0.0043,-0.0245,-0.0023,-0.0003,-0.0026,-0.0019 TUR_Marmara_Barcin_N:I0707,0.0108,0.0182,0.0005,-0.0341,0.0178,-0.0174,0.0006,-0.0051,0.0177,0.0465,0.0069,0.0109,-0.0146,0.0019,-0.0283,-0.0107,0.0175,0.0009,0.0103,-0.0022,-0.0096,0.0078,-0.007,0.0013,-0.0014 IRN_Ganj_Dareh_N:I1290,0.0032,0.0068,-0.0402,-0.0037,-0.0394,0.0043,0.0075,-0.0023,-0.0411,-0.028,0.0019,-0.0038,0.0062,-0.0106,0.0192,0.0519,-0.0044,0.0068,0.0116,-0.0259,0.0115,-0.0178,-0.0137,-0.0344,0.024

If a user is unsure of how to go about doing any of this, he/she can contact us and we will walk him/her through it. Alternatively, the individual can send us his/her G25 coordinates and we will break down that person’s ancestral makeup, free of charge.

I am an Amhara from Ethiopia. Why are my 23andme and AncestryDNA test results different from each other? And why are they different from the results of your genome analysis?

AncestryDNA uses Sub-Saharan African reference samples that already bear some Eurasian admixture. Consequently, when modern Cushitic, Ethiosemitic and North Omotic-speaking individuals have their genomes compared against these admixed proxy samples, the Eurasian ancestry carried by these Afro-Asiatic speakers is under-reported and their Sub-Saharan African admixture is inflated. For example, if the admixed Nilotes of the Pastoral Iron Age are utilized as surrogates to infer ancient Nilo-Saharan ancestry, Afro-Asiatic speakers from the Horn will show elevated Pastoral Iron Age affinity and comparatively lower Eurasian affinity. This is because the PIA Nilotes already bear substantial Eurasian admixture, which was specifically derived from the earlier Cushites of the Pastoral Neolithic, whom they absorbed. However, if the newly-discovered, “pure” (i.e. relatively unadmixed) ancient Nilo-Saharan Kakapel_300BP sample is used in place of the admixed PIA Nilotic samples, the true, low level of Nilo-Saharan admixture and high level of Eurasian ancestry is revealed.

Regarding 23andme, this company has a practice of pre-assigning a customer’s ethnic group to a geographic region depending on where that person’s ethnic group is centered. Since Amhara individuals come from Ethiopia (a country located below the Sahara), 23andme reserves an exclusive “Ethiopian & Eritrean” ancestry category for its Amhara and other Abyssinian customer genomes. It then places this “Ethiopian & Eritrean” category under a broader “Northern East African” heading, which, in turn, it automatically pre-assigns to its “Sub-Saharan African” ancestry region. This classification system has the effect of giving the false impression that Amhara customers are almost 100% of Sub-Saharan African ancestry, when actually the opposite is nearer to the truth. That is, most Amhara individuals and other Ethiosemitic, Cushitic and North Omotic-speaking persons from the Horn are predominantly of Eurasian ancestry (over 70% on average), and much of that Eurasian ancestry was derived from North Africa.

Originally, 23andme’s ancestry analysis used to reflect this fact, albeit to an exaggerated degree. This can be seen on the lab’s ancestry chart below (bottom left), which estimates a 98% Eurasian ancestry for a Cushitic-speaking individual from Ethiopia. In the early 2010s, around the time when 23andme debuted its v2 chip, the company adopted a new policy of pre-assigning ancestry according to geographic provenance. This unfortunate decision has since led to the misrepresentation of the ancestral composition of numerous paying customers, including Afro-Asiatic-speaking individuals from the Horn, Nubians and Sudanese “Arabs” from the Nile Valley, North Africans, South Asian persons, and Ashkenazi Jews who settled in Europe.

Regarding Nubians and Sudanese “Arabs”, 23andme earmarks a “Sudanese” category for these individuals. The designation falls under its broader “Northern East African” heading, which it again pre-assigns to a “Sub-Saharan African” ancestry region. This makes little sense because most Nubian and Sudanese “Arab” persons also have a predominant Eurasian ancestry (over 70% on average), most of which is shared with neighboring Egyptians. 23andme’s geography-based scheme likewise distorts the ancestral makeup of North Africans. It allots Egyptian, Libyan and Maghrebi individuals their own categories, which it places under a broader “Western Asian & North African” ancestry region. This serves to obscure the indigenous North African Iberomaurusian ancestry which all North Africans carry (except Coptic Egyptians) and which differentiates them from populations in western Asia; it also deflates the minor Sub-Saharan African admixture borne by these individuals. Furthermore, 23andme lumps all persons from the Indian subcontinent under one large “Central & South Asian” category, irrespective of whether they are a Punjabi Sikh with ties to Iranians or an Orissa tribal member with elevated Ancestral South Indian (Onge-related) ancestry. Similarly, 23andme and AncestryDNA set aside an “Ashkenazi Jew” category under a broader “European” ancestry region. The MyHeritage and Family Tree DNA testing companies, in contrast, have a special “Jewish Diaspora” ancestry region for their Ashkenazi and Sephardic Jewish customers, which is separate from their “European” and “Middle Eastern” ancestry regions. Ultimately, both of these classification schemes conceal the reality that Ashkenazi and Sephardic Jews are actually intermediate between Levantine individuals (Lebanese) and southern European individuals (northern Italians).

So why, then, do we recommend that customers buy ancestry kits from AncestryDNA, 23andme and other genetic testing labs? Strictly for the genotype raw data that he/she will be able to access (as well as for the haplogroup information), not for the factually incorrect interpretation of that data. Using the Vahaduo Admixture JS program and official Global25 coordinates, an individual can analyse his/her own ancestral makeup with much greater exactitude than any commercial testing firm is presently able to offer.

Ancestry results of a Cushitic-speaking individual from Ethiopia, from the genetic testing company 23andme (v1 chip). Originally, 23andme used to report that Afro-Asiatic speakers from the Horn region bore a predominant Eurasian ancestry (typically >90%). In the early 2010s, after introducing its v2 chip, 23andme decided to automatically pre-assign all Horn of Africa genomes to its “Sub-Saharan African” ancestry category. This is despite the fact that most Cushitic, Ethiosemitic and North Omotic speakers actually carry a predominant Eurasian ancestry (over 70% on average).

Why are the results of your genome analysis different from the census?

To properly answer this question, we need to differentiate between biological ancestry and societal classification. Biologically, the Cushitic, Ethiosemitic and North Omotic-speaking populations of the Horn of Africa carry a predominant Eurasian ancestry. However, socially, these groups have been categorized in various ways.

In the colonial period, Cushitic, Ethiosemitic and North Omotic speakers were frequently included among the foreign “races” on the official non-African censuses (e.g. the 1957 Census of British Tanganyika and the 1962 Census of British Kenya; see here and here). Many censuses in the West nowadays adhere to geography-based categorization schemes (France is an exception since it has legally forbidden government from collecting any ethnicity information; other nations will probably eventually follow suit). Consequently, in the United States, Afro-Asiatic speakers from the Horn and northern Sudanese have been assigned a Sub-Saharan African census categorization. This is at odds with these groups’ earlier social classification and actual biological affinities. However, individuals from the Arab League states of Djibouti, Somalia and Sudan are presently being considered for inclusion in a potential new Middle East and North Africa or “MENA” category (USCB). In Britain, the situation is rather different because, as of the 2021 UK census, Somalis have the option to choose a separate ethnicity category labeled “Other ethnic group: Somali” (classification code #276; cf. ONS). Besides White Africans (Afrikaners, White Zimbabweans), they are the only residents from Sub-Saharan Africa who have been accorded this opportunity. Ethiopians, Eritreans and northern Sudanese, despite being closely related to Somalis, have not yet been granted any such option. Moreover, in the Arab world, these individuals are usually categorized as Arabs; particularly those from Arab League member states.

Needless to say, societal classification is a complex matter that, unlike biology, is often subject to change depending on sociopolitical circumstances. Our focus on this website is instead the biological ancestry and culture of the Puntites, their modern descendants, and affiliated populations.

What is the difference between a Vahaduo Single analysis and a Vahaduo Multi analysis?

A Vahaduo Single analysis is ideal for determining an individual’s precise ancestral composition. The program sifts through the Global25 datasheet’s thousands of samples to pinpoint the most optimal ancestries for that person (no hand-picking of proxies or guesswork required on the user’s part). Once an individual’s ancestral elements have been identified, a Vahaduo Multi analysis can then be conducted using the “best representatives” of each of those components (if, for instance, a Single analysis assigns an individual CHN_Huatuyuan_500BP at a frequency of 15% and VNM_N at a lower frequency of 5%, CHN_Huatuyuan_500BP is the “best representative” East Asian element to use in the subsequent Multi analysis). A Multi analysis will let a user more easily visualize and interpret the analysed data for multiple individuals, including average component percentages.

Vahaduo Single analysis of a southern Somali individual. Single analyses help identify a person’s exact ancestral composition. In this case, the Single analysis indicates that this southern Somali individual has a predominant Eurasian ancestry (~77%), which consists of majority West Eurasian elements and a minority East Eurasian element. This is a typical result for a Cushitic/Ethiosemitic/North Omotic-speaking individual from the Horn region.

Vahaduo Single analysis of a Tigray individual from Ethiopia. This Ethiosemitic-speaking Tigray Abyssinian has a similar overall ancestral makeup as the Cushitic-speaking southern Somali. However, certain differences can be observed. Notably, the Tigrayan individual has a higher percentage of the Levantine Natufian component, in line with the adoption of Semitic languages by the Abyssinians’ Cushitic-speaking Agaw ancestors. The southern Somali individual also has greater frequencies of European Steppe elements and a slightly higher admixture of East Asian elements.

Vahaduo Single analysis of an ancient Cushitic individual of the Pastoral Neolithic. This person shares the same basic ancestral composition as modern Cushitic, Ethiosemitic and North Omotic speakers, indicating that overall there has been genetic continuity between the Cushites of the Pastoral Neolithic and their modern descendants in Northeast Africa.

What is the difference between proximal genome ancestry and distal genome ancestry?

Proximal or near genetic models are useful for identifying pulses of recent gene flow. For Cushitic, Ethiosemitic and North Omotic-speaking individuals, the ancient Cushitic samples of the Pastoral Neolithic and the medieval Kulubnarti samples should appear as their main proximal components.

Distal or remote genetic models are useful for ascertaining an individual’s overall ancestral makeup since, in theory, they employ the “purest” (i.e., least-admixed) samples available. For Cushitic, Ethiosemitic and North Omotic-speaking individuals, the Eurasian ancestries described above should appear as their main distal components.

Why do you prefer Global25 over other genetic analysis technologies?

Global25 (G25) allows a user to quickly and easily cross-analyse the genome of any individual or population with those of 6000+ ancient and modern samples. Eurogenes, which developed the system, also regularly updates its G25 datasheets to include any new decent-coverage samples that have been published. When used in conjunction with the Vahaduo Admixture JS program, these factors serve to make Global25 the premier technology available today for free-form genomic analysis.

Genetic programs that analyse population structure, such as ADMIXTURE, are the next best tools around. However, they are inherently limited by the number of populations a user may compare at a time (the K= rarely exceed 20). Some genetic affinities that do exist may therefore be overlooked or under-reported.

Genetic calculators consist of a preset collection of Source populations against which a user can input and compare his/her genome data (e.g. the GEDmatch K13 calculator). These tools can be useful for getting a rough idea of a Target individual’s possible ancestral composition. However, genetic calculators are inherently limited by the fixed number of proxy samples they can accommodate at any given time. The accuracy of their output depends entirely on how well-chosen those pre-selected reference samples are vis-a-vis the Target individual being analysed.

Genetic modeling programs like qpAdm are even more restricted in terms of the amount of populations they can simultaneously analyse (two to five are the norm). As such, a user must have a firm idea beforehand as to which Source populations contributed to the gene pool of a given Target individual or group. This makes these programs vulnerable to confirmation bias. Furthermore, if an individual’s or group’s ancestral composition includes six or more contributing Source populations — which, as it turns out, is the case with all modern Afro-Asiatic speakers in the Horn and North Africa (except Coptic Egyptians) — these tools will fail to capture a significant portion of their ancestry. Hence, qpAdm and similar genetic modeling tools are not well-suited for discovering an individual’s or population’s ancestral makeup. They are instead adequate for confirming population affinities that Global25 or ADMIXTURE have already identified, as long as the contributing Source populations are few in number.

I am a longtime reader of your blog. You have regularly made predictions that turned out to be true. What is your secret?

Two words: multi-disciplinary analysis. Our working ethos is that a convergence in the findings of multiple scholars and thinkers from different fields — including our own private research — is much more likely to be correct than that of just one authority from a single discipline. When we have stuck to this guiding principle, it has proven to be remarkably accurate.

I like this website. How can I financially contribute?

This website is a free resource. We don’t presently need any financial contributions (though we appreciate the thought).

If you really want to help, kindly share our links with others who might be interested.

How can I contact you?

We can be reached through the comment section below or on any of the essays on this website.

Advertisement