African American DNA: Ancient Origins, Source Regions and Limits

African American genomes carry a history that was, for most of its recent length, written down by other people. Between the early sixteenth century and the 1860s, some 12.5 million Africans were taken across the Atlantic in the slave trade, and roughly 400,000 of them were landed in what became the United States. Their descendants are the people whose genomes this post is about, and the genetic literature on that history now names the source regions, the proportions and the sex bias plainly.

One rule before the numbers. Ancestry is not identity. Being African American is a matter of history, community, culture and family, and no admixture proportion says anything about who someone is. What follows describes the populations these genomes resemble and where the evidence stops.

The short answer: African American genomes average around three quarters African ancestry, drawn mainly from West and West Central Africa (Senegambia, the Bight of Benin, the Bight of Biafra and the Congo-Angola coast), about a quarter European ancestry, mostly northwestern European and carried disproportionately through men, and a small Indigenous American share near one percent. The African side has almost no ancient genomes to compare against, so "ancient origins" for the African stream means deep population structure and the Bantu expansion rather than named ancient individuals.

The source regions#

Micheletti et al. (2020) analysed more than 50,000 people across the Americas and compared their genomes with African reference populations and with the shipping records of the Trans-Atlantic Slave Trade Database. For the United States, the African ancestry traces mainly to four regions: Senegambia (Senegal, Gambia and Guinea), the Bight of Benin (Ghana, Togo, Benin and southwestern Nigeria), the Bight of Biafra (southeastern Nigeria and Cameroon) and West Central Africa (Congo and Angola). Nigerian-related ancestry is more common than the direct shipping records to North America would predict, which the authors attribute in part to the trade of enslaved people from the Caribbean into the United States; Senegambian-related ancestry is somewhat under-represented relative to the records.

Bryc et al. (2015), using 23andMe data on about 5,000 self-identified African Americans, found an average of 73 percent African, 24 percent European and 0.8 percent Indigenous American ancestry. The African share ranges from well under half to nearly all of the genome, and it is on average higher in the South, especially South Carolina and Georgia, and lower in the Northeast and West. The Mexican and Latino post covers the same three streams in a different mixture.

The sex bias#

The literature documents a strong asymmetry in how the European and African contributions entered. Surveys of African American men find European Y-chromosome lineages, mostly R1b, in roughly a quarter to a third, while European mitochondrial lineages appear in under a tenth. Lind et al. (2007) titled their paper on exactly this point. Micheletti et al. (2020) estimated that African women contributed more to the gene pool of the Americas than African men, with the bias strongest in Latin America and milder, but present, in the United States, alongside a heavily male-biased European contribution. The history behind those numbers is the sexual exploitation of enslaved women by the men who owned them.

What "ancient origins" can mean for the African stream#

The honest answer starts with what is missing: ancient DNA from West Africa barely exists. The main exception is Shum Laka in Cameroon, where Lipson et al. (2020) sequenced four children buried around 8,000 and 3,000 years ago. Those individuals were not ancestral to present-day West Africans: they belonged to a lineage related to Central African foragers such as the Aka and Mbuti, with an additional deeply divergent component, and the population that gave rise to Bantu speakers and to most present-day West Africans is not represented by them. Other ancient sub-Saharan genomes, from Malawi, Tanzania, Kenya and South Africa, lie thousands of kilometres from the slave trade's source regions.

What the genomes can say about the African stream is structural. The populations of West and West Central Africa descend from a deep West African lineage that diverged from eastern and southern African lineages tens of thousands of years ago. Within that, the Bantu expansion that began near the Nigeria-Cameroon border around 3000 BC and reached the Congo basin and southern Africa over the next three thousand years is the largest demographic event, and it is why the Congo-Angola component of African American ancestry is close to, but distinguishable from, the Bight of Biafra component. Senegambian ancestry sits on a separate gradient shaped by the Sahel, with a small North African-related contribution that the Nigerian and Angolan components lack.

The European and Indigenous American sides are, by contrast, among the best-sampled ancestries in the world. The European ancestry of African Americans comes mostly from the British Isles, traced through the ancient genomes described in the English, Irish and Spanish and Portuguese posts. The Indigenous American share connects to the Anzick-1 lineage described in the Mexican and Latino post.

A dated timeline#

• Before 50000 BC. The West African lineage separates from other African lineages.
• About 6000 BC. Shum Laka's earlier burials in Cameroon; forager-related people, not ancestral to later West Africans.
• About 3000 BC. The Bantu expansion begins from the Nigeria-Cameroon border region.
• About 1000 BC to 500 AD. Bantu-speaking farmers reach the Congo basin; iron working spreads across West Africa.
• About 500 to 1500 AD. The Sahel empires of Ghana, Mali and Songhai; the Yoruba, Edo and Igbo polities; the Kongo kingdom.
• 1526 to 1808. The transatlantic trade to North America, most intense in the eighteenth century, landing people mainly at Charleston and the Chesapeake.
• 1808 to 1865. The internal trade moves people from the upper South to the cotton states.
• 1910 to 1970. The Great Migration carries families to northern and western cities, which is why the regional gradient in ancestry proportions is now blurred.

What the Y-DNA and mitochondrial DNA add#

Paternal lineages among African American men are dominated by E1b1a (E-M2), the majority haplogroup across West and West Central Africa; surveys place it near two thirds of African American men. European lineages, mostly R1b with smaller shares of I, J and G, account for roughly a quarter to a third. Other African lineages appear at low frequency, and Indigenous American Q is rare.

Maternal lineages are overwhelmingly African. Branches of haplogroup L, especially L2, L3 and L1, account for around 90 percent of African American mitochondrial lines. European lineages appear in under a tenth, and Indigenous American lineages (A2, B2, C1, D1) at a few percent, more than the autosomal share would suggest and consistent with Indigenous women marrying into African American families in the colonial period. No haplogroup is an "African American gene"; each lineage is one ancestor per generation.

Limitations#

• The African side has almost no ancient genomes. Shum Laka is the main exception and is not ancestral to most West Africans. A model against ancient sources describes the European and Indigenous sides in detail and the African side in outline.
• Modern reference sampling within Africa is uneven. Nigeria and Gambia are well represented; Angola, Congo, Togo and Benin far less so.
• Averages are not people. The 73 percent figure describes one 2015 sample. Individual genomes span the whole range.
• The records and the genomes disagree in places. The shipping database and the genetic estimates differ on Senegambia and Nigeria, and the reason is only partly understood.
• Nothing here is identity. A European share does not make anyone less African American, and no proportion certifies belonging to any community.

Frequently asked questions about African American DNA#

Where in Africa do African Americans come from?#

Mainly West and West Central Africa. Micheletti et al. (2020) traced the African ancestry of people in the United States to Senegambia, the Bight of Benin, the Bight of Biafra and the Congo-Angola coast, with Nigerian-related ancestry more common than the direct shipping records predict. A single genome usually carries ancestry from several of these regions.

How much European ancestry do African Americans have?#

On average about a quarter. Bryc et al. (2015) found a mean of 24 percent European ancestry in about 5,000 self-identified African Americans, with individuals ranging from almost none to a majority. It entered mostly through European men, which is why European Y-chromosome lineages are far more common than European mitochondrial lineages.

Are there ancient genomes from West Africa?#

Almost none. Shum Laka in Cameroon, sequenced by Lipson et al. (2020), is the main exception, and its four individuals belonged to a Central African forager-related lineage that is not ancestral to most present-day West Africans. Other ancient sub-Saharan genomes come from East and southern Africa. Ancient Matches on the African side are therefore sparse compared with the European side.

Can DNA tell which African people or tribe my family came from?#

Not precisely. Genome-wide analysis can place African ancestry in a broad region such as Senegambia, southern Nigeria or Angola, and often which region contributes most. It cannot name a specific people, because West African populations are closely related, reference sampling is uneven, and most genomes carry ancestry from several regions at once.

How to model African American ancestry with qpAdm and Global25#

A qpAdm model for an African American genome tests the three continental streams directly, and the catalog supplies a source for each: Sub Saharan African for the African side, Native American (15000 - 1500 BC) for the Indigenous side, and for the European side either the distal pair Anatolian Neolithic Farmer (8500 - 6000 BC) plus Western Steppe Herder (5000 - 2800 BC) or the proximal Iberian (800 - 50 BC), which is the catalog's nearest western European Iron Age reference. The model returns each weight with a standard error and a p-value, so a real Indigenous American share shows as a weight significantly above zero rather than a rounding error.

One thing should be said plainly. The African stream is represented by a single broad source in the catalog today. The model can measure the three continental weights with real error bars, but it cannot resolve whether the African ancestry is Senegambian, Nigerian or Angolan, because no ancient genome from those regions exists to serve as a source. The Global25 analysis, whose worldwide reference set includes modern West and West Central African populations by region, is the complementary step for that question, and Ancient Matches reports shared segments with individual ancient genomes, densest on the European side. All three are part of the ancestry service.

Sources and further reading#

1. Micheletti, S. J. et al. (2020). Genetic consequences of the transatlantic slave trade in the Americas. American Journal of Human Genetics, 107, 265 to 277.
2. Bryc, K. et al. (2015). The genetic ancestry of African Americans, Latinos, and European Americans across the United States. American Journal of Human Genetics, 96, 37 to 53.
3. Lipson, M. et al. (2020). Ancient West African foragers in the context of African population history. Nature, 577, 665 to 670.
4. Lind, J. M. et al. (2007). Elevated male European and female African contributions to the genomes of African American individuals. Human Genetics, 120, 713 to 722.

Editorial note: the hero artwork in this article was generated with AI as conceptual illustration. It does not reproduce a scientific figure, an ancient individual, a real site or a measured migration route.