The first effort to map variations in the human DNA code reveals patterns in three ancestral populations — European, African and Chinese — that go back thousands of years, scientists reported Thursday. They said the trend could lead to medical treatments tailored for your specific genetic background, but they also worried that it could spawn a 21st-century style of racial discrimination.
The work was unveiled here at the annual meeting of the American Association for the Advancement of Science, and published in Friday's issue of Science, the AAAS' weekly journal. Science's editor-in-chief, Donald Kennedy, said the data "will provide an invaluable resource for genetic research to improve human health."
Once researchers put the database to use, it shouldn't take long to find out whether the dream of personalized genetic medicine can become a reality, said David Cox, chief scientific officer at California-based Perlegen Sciences and a principal author of the study.
"There's no reason why we shouldn't know in the next couple of years how useful this concept of using genetics to improve existing therapies is going to be," Cox told reporters. "So it's an important day for us, because we've started the clock."
The study focused on individual variations in the millions of chemical "letters" that are strung together in our DNA code. These variations are called single-nucleotide polymorphisms, or SNPs ("snips"). If our DNA serves as a kind of "source code" for our cells, then SNPs can be seen as the variations that code for one individual rather than another.
Geneticists have found that some patterns of SNPs can be linked to particular traits, including predisposition to diseases ranging from cancer to obesity. Thus, identifying a particular pattern of SNP variations could suggest a particular treatment to ward off illness.
The researchers behind the latest study analyzed 1.58 million SNPs for 71 Americans of European, African and Han Chinese ancestry. To streamline their work, they used mathematical formulas to determine the linkages between SNPs in particular blocks of the human genome. Small sections of the DNA code containing a SNP could be used to predict larger, related chunks of the genetic sequence.
For all three groups, variations of particular SNPs were found in most cases: 93.5 percent for African-Americans, 81.1 percent for European-Americans and 73.6 percent for Chinese-Americans. That told the researchers that the African-Americans had the most genetic diversity. But the researchers said their sample group of 71 individuals was far too small to make generalizations about whole populations.
Perlegen's Cox stressed that the research shouldn't be used in the debate over racial differences.
"People cannot use our data to say, 'See, I told you there are races,' or, 'See, I told you there aren't races," he said. "Nor can they say, 'See, the differences are more important than the similarities.' But these data will be useful for starting to address such questions as which kinds of medical treatments should be used, based on physiological differences caused by genetic variations."
The Food and Drug Administration already takes in data about ethnicity to profile how experimental medicines affect different segments of the population, noted Lawrence Lesko, the agency's director of clinical pharmacology and biopharmaceutics. The new genetic database should make that process more precise, he said.
"The way this information is currently gathered is like telling time by looking at a sundial rather than looking at a Rolex watch," he said.
In a Science commentary on the research, Troy Duster of New York University's Institute for the History of the Production of Knowledge said that researchers would have to "climb back on the tightrope to address racial disparities in health" and avoid reinforcing old concepts of race.
"One should be very alert and very careful," he warned.
Cox's colleagues in the research are David Hinds, Laura Stuve, Geoffrey Nilsen, Dennis Ballinger and Kelly Frazer, all from Perlegen; Eran Halperin of the International Computer Science Institute in Berkeley, Calif.; and Eleazar Eskin of the University of California at San Diego. The SNP map is being made available to other researchers through as well as the .
An even more detailed map of SNP variations is expected later this year from the International HapMap Project.
In the future, researchers looking for individualized cures could take advantage of the fact that SNPs are interrelated in mapped "neighborhoods." For example, a quick DNA test could determine that for your genetic profile, one drug would be more effective than another for treating hypertension or warding off Alzheimer's disease.
"Any large global-based study certainly can be useful for those of us who are much more focused on how we look at variations in individual genes," said Leroy Hood, president of the Seattle-based Institute for Systems Biology, which hopes to develop individualized medical tests and treatments.