January 8, 2011

Exploring the ancestry of African-Americans

Preliminary note: in this post African-American means people of African descent living in America (aka the Americas), not just people from the USA.

The authors acknowledge that they find difficulties in tracking regional African ancestry because of the limited genetic data available for this continent, with some important countries in the Atlantic slave trade, like Ivory Coast or Ghana, being still largely not researched. The situation may be even worse for Y-DNA and autosomal markers.

I'd say that this sad situation is product of several factors: first of all lack of interest by Western or Asian researchers, then unequal access to the DNA pools of the continent, often for political reasons (this happens elsewhere, specially in parts of Asia but also in France), other factors are the huge size and diversity of Africa and its population and the relative lack of obvious structure, with many lineages scattered across the continent. This last element actually seems to demand greater research and finer-grained haplotyping but this only comes in small drops. 

Inter-continental admixture, the well known sex-bias

A simpler analysis can be done in regards to inter-continental admixture (fig. 1):

Fig. 1 - Ancestry of African-Americas: (a) mtDNA, (b) Y-DNA and (c) autosomal DNA
We can see in these maps, for example, that Afro-Brazilians are largely of European and Native American ancestry (c), even if African descent is still dominant.  We can also see that European ancestry is overwhelmingly of male lineages (b), while Native American one instead is almost only of female origin (a). 

This seems quite peculiar of Afro-Brazilians, with no other American population displaying such pattern. Although it may be arguable in the case of Afro-Caribbeans. The population displaying most European female ancestry are African-Americans from Philadelphia.

Notice that minor European and Native American components are present in the autosomal samples of pure West Africans, indicating that these minor amounts are actually imprecisions of the clustering method. 

In the reverse direction, there is a map (fig. S1) in the supplementary materials that shows important African and Native American ancestry among white Brazilians, again showing sex-bias in the admixture, with nearly all being attributable to female ancestry. Some admixture is also apparent in white Philadelphians but it is much smaller.

African regional ancestry estimates

But the greatest effort of the authors of this paper is trying to unravel the African regional ancestry of these populations. For this they had to resort to mitochondrial DNA, which is admittedly just part of the whole picture and is often not sufficiently well structured by regions anyhow:

Fig. 3 regional African ancestry estimates (mtDNA)
The impression we get is that SE African (Mozambique) and Gabonese ancestry is restricted to South America. Instead Cameroon ancestry is concentrated towards the North of the new continent. Angolan and West African ancestry are widespread, however there is no detected Angolan ancestry in the former Spanish colonies. 

The authors argue some of these differences based on the chronology of the Atlantic slave trade, that in its first phase almost exclusively preyed in Westernmost Africa and Angola, while looking for other sources West-Central Africa and Mozambique in its late phases.

West African insular peoples

A map available in the supplemental material, fig. S6, also addresses the matrilineal ancestry of the populations of Cape Verde and São Tomé, two insular areas uninhabited before the colonial period and settled essentially with slaves from the continent (though Cape Verdeans  specially also have some important European admixture, less important probably in São Tomé). Cape Verde African ancestry is exclusively from Westernmost Africa, including Mauritania, instead São Tomé shows a mixture of Westernmost Africa and Central Africa (Gabon, Angola), to the exclusion of the Nigeria-Cameroon area.


  1. In fairness to our genetic researchers, significant portions of Africa have been immersed in wars, civil strife, political unrest, natural disasters and dire poverty for large chunks of the time in which genomic analysis has been possible, making this research difficult and a fairly low priority in the greater scheme of things.

  2. I understand the difficulties but there is also lack of interest. Most African states have been fairly stable in the last decade, which is the one when nearly all genetic research has been made. If my memory is correct, the most conflictive states are those of the Central-Eastern Sahel (Niger, Chad, Sudan, Eritrea and Somalia), Ivory Coast and DR Congo. All the rest of Africa has been quite stable for the last many years.

    I believe that the lack of people of African ancestry (either Africans or African-Americans) among researchers have played a role. Almost uniformly geneticists, are of European or Asian ancestry. And, while it is worth recognizing the important efforts of some of these researchers in Africa, the continent remains up to this day quite shallowly researched, specially if we compare with Eurasia and America.

    I do not think it is low priority at all. Even from a purely Eurasian viewpoint, understanding African genetics is crucial to understand the genetics of the rest of the World. Additionally, there is a major fraction of the World, in Africa but also in America, who have more direct roots in this continent and who could benefit intellectually or even medically from such research.


Please, be reasonably respectful when making comments. I do not tolerate in particular sexism, racism nor homophobia. Personal attacks, manipulation and trolling are also very much unwelcome here.The author reserves the right to delete any abusive comment.

Preliminary comment moderation is... ON (your comment may take some time, maybe days or weeks to appear).