View taken July 8, 2010, of a mannequin of Tautavel Man presented at the prehistoric museum in Tautavel in France

Dear Professor Gates:

I read your column “Do Most Whites Have Traces of African DNA, as I Do?” with interest. The questioner in that column was wondering whether the traces of African ancestry in her DNA test entered her family tree recently (perhaps during slavery) or in ancient times (remnants from all of humankind’s origins in Africa).

You told her, “Although some DNA tests probe into deeper ancestry, back thousands of years, the autosomal test that you took from 23andMe generally shows more recent ancestry, quite reliably over the last 100 to 200 years—in your case, since the time of slavery, when this ‘admixing’ most likely occurred. Given this information, the African DNA identified in your test is not from ancient ancestors (after all, 50,000 or so years ago, all of our ancestors were Africans, but that ancient DNA has largely disappeared) but, rather, from more recent ancestors … ”

I, too, am white, and a DNA test showed that I have 0.1 percent sub-Saharan African ancestry. I assume some of my ancestors (slave owners, unfortunately) had children with their slaves, who then “passed” as white.

However, here is where I get confused: My DNA test also showed that I have 2.6 percent Neanderthal ancestry. The testing service 23andMe states on the Web page explaining my results that the Neanderthal DNA goes back thousands of years: “Genetic evidence suggest[s] that they interbred [with modern humans] and although Neanderthals disappeared about 40,000 years ago, traces of their DNA—between 1 percent and 4 percent—are found in all modern humans outside of Africa.”  


Doesn’t this conflict with the above explanation of the autosomal results? How could the African test results only come from the last couple of centuries, while the Neanderthal results go back thousands of years? —Lorraine Gardner

I can see how the difference in timescale can be confusing. To help explain why the results are not in conflict, I went straight to the source. Kasia Bryc, a population geneticist with 23andMe, emailed this response to me:

DNA carries a lot of information. The challenge is whether there is a strong enough signal in the DNA to answer the question you wish to investigate, and that is determined by a lot of factors.

To estimate ancestry for Ancestry Composition [the report where you saw your African results], we use reference data of individuals whose ancestry we know, in order to figure out the ancestry of individuals whose ancestry we don’t know. Using reference data allows us to build an algorithm to disentangle ancestry from different parts of the world, but this is actually quite a hard problem.

The challenge is that present-day populations across the globe are genetically very similar to each other. Most genetic variation is shared around the world, across all populations, so our algorithm relies on rather subtle differences in patterns of genetic variation to estimate ancestry from different population and regions.

23andMe’s Neanderthal estimates come from a separate analysis from Ancestry Composition, using a different method that takes advantage of having DNA that is 40,000 years old from Neanderthal individuals.

The key difference between estimating African or European ancestry and estimating Neanderthal ancestry is that, although Neanderthal introgression [i.e., interbreeding with modern humans] occurred tens of thousands of years ago, Neanderthals as a population had been separated from modern humans for a long time before they mixed together, probably for at least 400,000 years, which translates into a lot of generations! (Plus, Neanderthals are likely to have had a comparatively smaller effective population size.)

What that means is that, over that time, Neanderthals became genetically differentiated from modern humans through genetic drift. Not only did Neanderthals acquire new mutations, but the patterns of genetic variation also drifted to be quite distinct from those of modern humans. As a result, it is relatively easier to pick up on Neanderthal ancestry, because the Neanderthals themselves were quite distinct—even though the mixture between Neanderthals and modern humans happened a long time ago.


Bryc also cautioned against reading much into your particular African DNA result, since, at 0.1 percent, it is quite small. It’s hard to know whether the test is picking up on recent ancestry, ancestry much farther back or a genetic pattern by chance. She addressed a similar scenario in a previous column, “How Long Ago Did African Ancestry Enter My Family Tree?” Reading it might shed some light on your own situation.

Henry Louis Gates Jr. is the Alphonse Fletcher University Professor and founding director of the Hutchins Center for African and African American Research at Harvard University. He is also chairman of The Root. Follow him on Twitter and Facebook.

Send your questions about tracing your own roots to