Discover - November 2009

dad called with news. “Guess what?” he said eagerly. “I got more of my DNA results. I’m mostly from eastern England where the Puritans are from. I belong to haplogroup I2a.” This didn’t surprise me; I2a is a branch of haplogroup I, to which 20 percent of all Europeans belong. Even my father’s speci;c branch, haplogroup I2, is pretty common.

Suddenly I felt like a child again—under the covers with my pajamas on, listening to a bedtime story—as my father told me about our migration history. “You see, Boonsri, the I2 people are the mammoth-eaters portrayed in The Clan of the Cave Bear,” he told me. People who belong to the I2a subgroup are probably from the

The next time I get the question
people always ask—Where
are you from?—I’m not about to
say, “I’m from haplogroup D4.”

Balkans and Sardinia; judging from the accumulation of mutations, this group split from haplogroup I about 25,000 years ago. My father must have descended from a pocket of people in Eastern Europe that migrated to England, perhaps along with the Roman army. The Puritan Dickinsons then moved to New England.

My dad said the report con;rmed everything he had already known about our genealogy, particularly the English part. After his father passed away, he discovered his family is related to Emily Dickinson. The connection is not direct, since she did not have any children, but her grandfather’s brother was an ancestor of mine.

23andMe: Scanning for Genome-Wide Associations

Much of the recent hype in ancestry testing has focused on genome-wide association tests, which can rapidly scan hundreds of thousands of SNPs. So for my third sample I ordered the kit for such a test from 23andMe ($399, 23andme.com). The company examines more than 550,000 SNPs and looks for associations to predict a person’s risk of common diseases, but it also roots out ancestral information. I decided to see this process live in the lab.

I drove to the top of a cliff in San Diego and stood before the daunting yet delicate glass structure that houses Illumina, the lab to which 23andMe had outsourced my sample. The pristine building is the size of a football ;eld; re;ections of the bright California sun nearly blinded me as I entered. Inside, I toured the state-of-the-art DNA sequencing facility. Technicians were dressed in white lab coats, shower caps, and gloves as they tested mailed-in samples for markers in all parts of the genome—not just the mtDNA and not just the Y chromosome.

When I returned home and logged into my 23andMe account, my data popped up in a Facebook-like page that listed my risk of various diseases as well as a report of my ancestry. “Friends” who shared their genomes with me were listed along the right side of the page. The report added some details about my place in haplogroup D: I belong to subgroup D4a. “Your maternal ancestor was probably from Siberia and northern China,” Joanna Mountain, senior director of research at 23andMe, explained. People in subgroup D4a have a gene for longevity, consistent with what I know about my grandmothers and great grandmothers, who all lived well into their nineties.

23andMe provided a map to show where my ancestors came from, but the results could not properly represent my mixed ethnicity. The company’s computer analysis correctly determined that roughly half my DNA came from Europe and half came from East Asia. Then, drawing a conclusion that only a computer could draw, it split the difference and indicated I was most genetically similar to people halfway between the two, in the middle of Central Asia.

A big part of the problem is the dif;culty of cross-checking
the results. “Your ancestry results totally depend on the genetic
database the company has collected so far,” says Deborah
Bolnick, a biological anthropologist at the University of Texas at
Austin. “Obviously, not everyone in the world has had
their DNA tested, and so companies must rely on their
own proprietary databases.”
In the end, 23andMe’s whole-genome approach
again mostly repeated details I already knew about
my family and added nothing to my understanding of
our recent history of migration. The next time I get the
question people always ask when they meet me—

Where are you from?—I’m not about to say, “I’m from haplogroup D4.” It will be a long time before ancestry companies can provide truly meaningful details at that level.

What Comes Next?

The most exciting thing in ancestry testing is happening not in the commercial market but in the research labs. At Harvard University, geneticist David Reich is applying a technique called admixture mapping to study the history of people of mixed descent, analyzing stretches of DNA to see where they come from and when mixing ;rst occurred. He has compared the genomes of indigenous populations in Africa and Latin America with the genomes of similar peoples who migrated and intermarried with other populations during the past several hundred years.

Reich ;nds that African Americans have patterns of genetic variation unmistakably different from those of the people in their ancestral country; Latin Americans have also diverged from the indigenous people of their native countries. These results show how the genome can record migrations on timescales of hundreds of years, not just thousands.

George Church, another Harvard geneticist, is working to speed up the sequencing and analysis of a person’s entire genome. Once that becomes available to the general population at an affordable price, he says, geneticists will be able to track your DNA letter by letter across generations to ;nd out whether a rare trait was inherited from your mother or from your father or acquired through a unique mutation. At that point, ancestry testing will become powerful and family-speci;c.

If only my results had been as impressive. The expanded sense of identity that I got while researching this article came mostly the old-fashioned way. I trekked cross-country to visit 23andMe, rode the bus to meet researchers at Penn State, and had new kinds of conversations with my father. Recently he told me that my great-grandfather had died from epilepsy at a young age, ;lling in one more piece of my family history.

Clearly I had hoped for too much from the ancestry tests.

In the future, DNA databases will expand and the new-style genealogists will get better at interpreting our genome. But the events that de;ne who we are will continue to unspool as they always have—through experience and serendipity.