BY MARK SHWARTZ
Scientists have long recognized that, despite physical differences, all human populations are genetically similar to one another. But a new study in the journal Science concludes that populations from different parts of the world share even more genetic similarities than had previously been assumed.
At the same time, researchers found that tiny differences in DNA can provide enough information to identify the geographic ancestry of individual men and women.
The results of the Dec. 20 Science study -- the largest of its kind to date -- have implications for understanding ancient human migrations and for resolving the ongoing debate about the use of ancestry information in medical research, said Marcus W. Feldman, the Burnet C. and Mildred Finley Wohlford Professor in Stanford's School of Humanities and Sciences. Feldman led the international study, which included scientists from the United States, France and Russia.
"Different populations experience different rates of various diseases," Feldman explained. "To determine if someone is genetically susceptible to a particular disease, doctors will sometimes ask a patient, 'What is your ancestry?' But recent studies have raised the question of whether self-reported ancestry is a useful diagnostic tool, or whether it should be abandoned in favor of genetic testing."
The authors concluded that a patient interview can, in fact, provide a useful, less invasive alternative for assessing individual disease risks.
In their study, Feldman and his colleagues analyzed DNA samples obtained from 1,056 people from 52 populations in five major geographic regions of the world: Africa, Eurasia (Europe, the Middle East, Central and South Asia), East Asia, Oceania and the Americas.
To identify specific populations, the research team looked for "microsatellites" -- short segments of human DNA that occur in specific patterns, which are passed down from generation to generation. In the study, researchers analyzed 377 microsatellites that population biologists routinely used as genetic markers.
"Each microsatellite had between four and 32 distinct types," Feldman said. "Most were found in people from several continents, suggesting that only a tiny fraction of genetic traits are distinctive to specific populations. This means that visible differences between human groups -- such as skin color and skull shape -- result from differences in a very small proportion of genetic traits."
Another way to view this, he noted, is to remember that DNA is virtually identical in all human beings. Compare any two people, you'll find a DNA sequence that is about 99.9 percent identical.
"In the less than one percent of the genome where genetic differences among individuals exist, it might seem intuitive that two people from different regions are likely to have more differences than are two people from the same region," Feldman noted. "But this is not the case. About 94 percent of genetic differences are among individuals of the same populations."
The conclusion, he said, is that people from different lands have more genetic similarities than scientists previously thought.
Although populations are genetically quite similar, Feldman and his co-workers wanted to see if they could predict where an individual's ancestors came from through DNA analysis alone. To accomplish this, the research team first removed the labels from all 1,056 DNA samples used in the study.
"We took the labels off of all the individuals so we didn't know where they came from," Feldman explained. "Then we asked the question, 'Can we look at the DNA and detect where groups of individuals form clusters that are genetically related to one another?'"
The answer should be yes, Feldman predicted, because, while most genetic types are widely distributed geographically, the frequencies of these types vary around the world.
In the study, lead author Noah A. Rosenberg of the University of Southern California and co-author Jonathan K. Pritchard of the University of Chicago applied a powerful statistical technique that uses many independent genes to detect the geographic patterns of ancestry in samples from any species. When applied to people, the technique proved remarkably successful. The research team accurately pinpointed the ancestral continent of virtually every individual from Africa, East Asia, Oceania and the Americas.
People from Eurasia -- which includes Europe, the Middle East, Central and South Asia -- were among the most difficult to assign ancestries, Feldman noted. "A complex history of migrations, conquests and trade over the past few thousand years is likely to be the cause for this difficulty," he said. An exception were the Basques of Spain -- a geographically and linguistically isolated population that was genetically distinguishable from other European groups.
The Science paper also supports recent genetic studies of human migration, confirming migratory patterns between Europe and West Asia, Europe and Central America and other continents as well. "By sampling genotypes from people from all parts of the world, geneticists have reconstructed the major features of our history: our ancient African origin, migrations out of Africa, movements and settlements throughout Eurasia and Oceania, and [the] peopling of the Americas," wrote Mary-Claire King and Arno Motulsky, both of the University of Washington, in an editorial accompanying the Science paper.
DNA analysis confirmed what most of study's 1,056 participants had said about their ancestry -- a finding that lends credence to the argument that an individual's own family history can be a useful way of determining his or her genetic predisposition to disease.
"Some members of the medical community argue that doctors simply shouldn't ask patients about their ancestry because it has no genetic meaning," Feldman noted, "but our study finds that self-reported ancestry and genetic ancestry are largely coincident, so patient interviews can be very useful. On the one hand, grouping patients by genetic similarities will benefit future studies that will scan the entire human genome for potential genetic causes of disease. On the other hand, self-reported ancestry makes it easier to get information on environmental factors, cultural differences and behaviors which may be important risk factors for certain diseases."
Other co-authors of the Science study are James L.
Weber of the Marshfield Medical Research Foundation in Wisconsin,
Howard M. Cann of the Center for the Study of Human Polymorphisms
in Paris, Kenneth K. Kidd of the Yale University School of Medicine
and Lev Zhivotovsky of the Russian Academy of Sciences in Moscow.
The research was supported by the National Institute of General
Medical Sciences. SR
Marcus W. Feldman
Stanford Report, January 8, 2003