Tracing African roots through DNA (cont.)

By John Simons, Fortune writer

Discovering the past

In mid-November I decided to make the same journey and delivered genetic samples to African Ancestry's offices. I took both of the tests the company offers - an analysis of mitochondrial DNA, passed to me from my mother and her mother, and so on, and a second test to examine my Y-chromosome, passed to me from my father and his father, etc. The cost: $550.

The sample collection was simple. For each exam, I rubbed my inner cheeks and gums for two minutes, collecting microscopic tissue samples on a swab, then sealed them into separate bar-coded envelopes. The specimens would first go to a Utah DNA processor called Sorenson Genomics, which sequences, digitizes and sends the DNA to Kittles in Chicago. Using proprietary software, Kittles tries to identify matches between his clients' DNA and those in his gene database. The whole thing takes four to six weeks.

What's interesting, and somewhat limiting, about the tests is that my mitochondrial DNA can only reveal information about my mother's mother's lineage. And my Y-chromosome analysis can only provide clues about my father's father's ancestry. That's all the genetic information I carry. What's left out, then, is information about my mother's father's DNA (passed through generations via the male line), and my father's mother's DNA, which is maternally inherited.

A few days before Christmas, I received my results in the mail. "The mitochondrial [i.e., maternal] DNA sequence that we determined from your sample shares ancestry with the Mende, Temne and Limba peoples in Sierra Leone today," read my official letter. Included in the envelope was a lengthy explanation of the genetic analysis, along with a graphic depicting portions of my DNA and a booklet offering information and sources for further research on specific ethnic groups.

"The Y-chromosome DNA markers," the letter continued, "share ancestry with people in two countries today: the Makua people of Mozambique and the Lissongo people in the Central African Republic. While these groups may differ socially and culturally, there are people within them who share a common genetic ancestry."

Kittles helped put my ancestry in context. About 30 percent of his clients have Sierra Leone matches, he says, which makes sense, as it's in the heart of the slave-trading region. Only about 5 percent of African slaves were from what is now Mozambique and Madagascar in the southeast, far from the main centers.

With this new information in hand, I now know more about my African ancestors than I do about many succeeding generations. The entire period between when my ancestors arrived (whenever that was) and the early 20th century is a mystery.

On my father's side, I know that my paternal great-grandfather was born in Brooklyn in 1910. His parents and their families, so I'm told, had lived in northern Virginia for generations. My mother's family hails from the Bahamas. Early details of their arrival in the U.S. are sketchy because they came as illegal immigrants and picked fruit in Florida. By the early 1940s, when my mother was born, they were living legally in Brooklyn.

My mother, visiting me for the holidays, couldn't contain her excitement upon hearing that she was related to the three main ethnic groups in Sierra Leone. She immediately commandeered my computer and began surfing Web sites for photos of the Mende, Limba and Temne people. "Oh, my God. She looks like Aunt Louise," she said, pointing to a photo of a Mende woman standing in a field with her children.

After a few more hours on the computer and a trip to the bookstore, my mother had a plausible theory about our Sierra Leone origins: Slaves from Sierra Leone were sought for their expertise in rice farming. Many of them ended up in the Carolinas and Georgia as the backbone of the region's rice industry. How did these possible ancestors get from the Carolinas to the Bahamas? Before the American Revolution, the Bahamas was a sparsely populated British outpost, but after the American victory in 1783, many plantation owners who remained British loyalists resettled in British colonies in the Caribbean. Between 1784 and 1789, the population of the Bahamas tripled to 12,000 people - three-quarters of them slaves.

My mother and I could not have made these kinds of connections, however tenuous, without the results from African Ancestry. But a number of scientists have questioned the interpretations the company makes (and are more than a little piqued that Kittles does not share his database). Last October, Bert Ely, a geneticist at the University of South Carolina, published a paper suggesting that African-American mitochondrial DNA has been mixed so much that, in many cases, it is impossible to find a match with a single ethnic group in Africa.

Ely plotted the DNA sequences of 170 African Americans against those of 3,725 people living in Africa. He found that most African Americans share lineage with three or more groups of Africans. He also found that some 40 percent of African Americans had no match with Africans in his database. In my case, it's not that my DNA doesn't match the Mende, according to critics, but that I may not have the full picture.

"You're limited by the size of the database," says Jonathan Marks, an anthropologist at the University of North Carolina. "How many other people did you match from other tribes that were not sampled?" The question is fair. Still, although the record is incomplete, I now have a link to my past I did not have before.

Pilar Ossorio, a professor of bioethics and law at the University of Wisconsin, questions the claims of specificity. "Just as an example," she says, "there are people in the Balkans who share the same mitochondrial DNA with people in Africa."

Kittles defends his research. His database is more than triple the size of Ely's, he notes; therefore, it has more ethnic variability and is more accurate. He grouses, "This is sour grapes that I didn't share my data with them."

African Ancestry's clients appear unfazed. On a Web forum, customers overwhelmingly side with the company. Dwainia Tullis's view is typical. Last August she discovered she shared ancestry with the Balanta and Fulani tribes of Guinea Bissau.

"I'm getting another cousin of mine involved to cover all sides of the family," says the 51-year-old hairdresser from California. And what has been the benefit? "I'm more whole now," she says simply.

For me, my family tree is still missing branches. Rather than one specific place of belonging, I've discovered I share DNA with people spanning the entire continent of Africa. In the end, I feel more African when I peer into the mirror - and, oddly, more American.


Jeff Hawkins hacks the human brain Top of page