The senior scientist and population geneticist at 23andMe, the California-based personal genomics and biotechnology company, talks to us about the first of its kind “genetic portrait,” ancestry composition reports, and surprises revealed by DNA testing.
By James Pylant
Copyright © 2016 | Posted 7 July 2016
Do not post or reprint without written permission
Previously a Postdoctoral Research Fellow at Harvard Medical School, Dr. Kasia Bryc received M.S. and Ph.D. degrees in Biometry at Cornell under Carlos Bustamante and a B.A. from Stanford in M&CS. “It’s a complexity, when what you think is true is not,” says Dr. Bryc about the results revealed by genetic testing. “DNA does not lie.”
James Pylant: You and your colleagues published the first of its kind “genetic portrait” of the United States. How was this project developed?
Kasia Bryc: This study, which was published last year—we refer to it as the genetic portrait of the U.S.—looks at the genetic ancestry of individuals who self-identify as European American, African American, and Latino or Hispanic. People are familiar with the consumer product at 23andMe, but often are less familiar with the research arm of the company. We invite customers to participate in research, and about 80% of individuals opt-in to participate. We invite them to answer questions about various traits of themselves, for example, we can ask: how do you self-identify? When we looked at our database, we had large sample sizes of individuals who self-identified as European American, African American, and Latino. (Obviously, other groups are represented in the database, but the sample sizes were smaller, so we didn’t look at that in this paper, but we hope to expand our research to other groups as sample sizes grow.)
The first question we asked involved looking at the genetic ancestry of individuals from these groups. When we look at your DNA, we estimate where in the world it might have come from, and we combine information from across your entire DNA. We can then say what percentage from different parts of the world. And then by looking at a lot of individuals sampled from different parts of the U.S., we can make inferences about ancestry—not just at the individual level, but also sets of individuals from different parts of the U.S.
JP: Let’s talk about 23andMe’s Ancestry Composition report. How do these results reflect recent ancestry vs. ancient ancestry?
KB: People like to talk about DNA in terms of time frames, but generally, the way that I explain it is, your DNA is tracing all your ancestors… and it very powerfully carries signals of all sorts of things. And so the question is, how do you extract the information you’re looking for from the DNA? You can ask different questions and get at different time frames based on what question you’re asking. We can see, for example, the Neanderthal DNA that’s carried by a lot of modern humans. That DNA was introduced presumably about 50,000 years ago into modern humans— but because it was so different genetically, we can actually see it pop up by comparing present day individuals to what the Neanderthal genome looks like. Which is how scientists were first able to infer that modern humans and Neanderthals mixed together, because the Neanderthal genome was quite distinct. For your ancestry composition, what we’re doing is looking at individuals, whose ancestry we know comes from different parts of the world, and we ask which populations your DNA looks most similar to. In the last 500 years since Columbus, there’s been a lot of migration, and so what we’re trying to do when asking this question is to use present-day individuals who are representative of what the DNA would’ve looked like around the world before those transcontinental migrations that occurred during the last 500 years.
JP: Sometimes there’s surprise or disappointment when admixture results don’t identify what we expect—such as Native American.
KB: If you’re looking at your parents, you’re inheriting 50 percent of your DNA from each of your parents. But one more generation back, you get only on average about a quarter of your DNA from each of your grandparents. What people don’t always appreciate is you don’t necessarily get exactly 25 percent of your DNA from each grandparent: what you inherit is actually random. So you might be slightly more related to one grandparent than another. Likewise, if you’re looking at going back five or more generations, then what DNA you inherited from a particular ancestor continues to be very random and enough generations back, it could be that you happened to get no DNA from a particular ancestor. So, for example, you might have had a Native American ancestor more than five generations ago from whom you didn’t inherit any DNA. This is still your genealogical ancestor, but they didn’t contribute DNA to you. That might be what happened, in cases where someone has been told by their family lore that, “You had a Native American family member,” but where they may not carry Native American DNA. Sometimes that’s the case because it’s too many generations back and it sort of “fell out” due to the randomness of what you inherit. For recent ancestors—your parents, your grandparents, your great-grandparents—it’s a zero percent chance that you wouldn’t get DNA from them, but when you start going enough generations back, it does happen—and you don’t inherit any DNA so you can lose those signals of an ancestry.
JP: Ethnicity reports can also differ according to the testing company, too.
KB: Yes, there are differences between the companies due to the different methodologies. In on our research, we confirm that our ancestry admixture estimates are quite accurate, and have validated our results in many different ways. For example, we saw that a lot of African Americans don’t carry much Native American ancestry, but mainly African Americans carry small fractions of Native American ancestry, and we’re able to confirm that the signal we’re picking is really evidence of Native American ancestry, even when you get down to that percent or half-percent level; it’s really is very accurate.
JP: What about the percentages of white Americans, like myself, who have African ancestry?
KB: One of the things that had been seen anecdotally, but not in a systematic way in European Americans until our paper, is that people that identify as European American often carry small amounts of African ancestry. In our paper we show that this “hidden” African ancestry occurs most commonly in individuals who are living in the South. We found about 10 percent of individuals who are European Americans—or people who identify as white—living in the South carry small amounts of African ancestry. Most of the time, we’re looking at small percentages around one or two percent African ancestry. Something like three-and-a-half percent—at least according to our estimates—of European Americans have one percent or more of African ancestry. We have maps of states where the proportion of European Americans with African ancestry is very striking, and is correlated to the population density of African Americans living in each state. I think that just shows what happens when people are living in the same place.
JP: DNA as a research tool opens doors for African American genealogy.
KB: DNA may reveal things like the family stories which have been lost or forgotten. That’s especially poignant for African Americans trying to research into their family history, and the paper records are just not there, and as you know, that was partly deliberate. DNA is so very powerful for African Americans, or others who know little about their family history or their roots.
JP: The results of DNA testing can be surprising, even to experienced genealogists.
KB: We find that customers often know or have some idea about part of their family tree. You may not be able to trace back all of your lineages, but with DNA, the signal of your ancestry is there. For example, people may see a hint of Scandinavian ancestry, and they were not expecting that. In my case, I’m Polish—born in Poland. I’ve been to my grandmother’s and great-grandmother’s and great-great-grandmother’s graves in the same small town in Poland. But in my results, I saw a small amount of Balkan DNA. I was completely surprised. We’re Polish! I brought this up with my mom, and she said, “Oh, did you know your father’s grandfather wasn’t from Poland?” And I said, “No, I didn’t know! You never mentioned that, in fact.” So for me it’s the power of being able to see the DNA, and to start the conversations with your family members. I would’ve never had that conversation with my mom if it hadn’t been for the DNA.