Scientists have shown that they can use genomic analysis to pinpoint a person’s geographical origins to within just a few hundred kilometers. Besides offering possibilities for the testing of genetic ancestry, the research could also have important implications for understanding the role of genes in complex diseases and other genomic-based health studies.

By plotting the differences between genetic variations of 3,000 Europeans in a two-dimensional grid, the researchers were able to reveal a pattern that looks remarkably like Europe. The scientists included researchers from Cornell University; the University of California, Los Angeles (UCLA); the University of Chicago; and the University of Lausanne, in Switzerland. The findings appear in this week’s issue of Nature.

Others have recently published similar research, in Current Biology, says John Novembre, a coauthor of the Nature paper and an assistant professor at UCLA. But the latest study goes further, by using algorithms to try to predict a person’s geographical origin based purely on his or her genetic variations, with a high degree of accuracy. The scientists were even able to reveal patterns of origin distinguishing French-, German-, and Italian-speaking groups within Switzerland.

In many respects, the results are not at all surprising, says Michael Krawczak of the Institute for Medical Informatics and Statistics at the Christian-Albrechts University of Kiel, in Germany, who took part in the Current Biology study. It was well established that the farther apart two people’s origins, the more different their genes will be, he says. “But it had never been shown before at a genome-wide level.”

One of the reasons that this is now possible is the plummeting cost of genotyping, says Novembre. The Affymetrix GeneChip measures 500,000 single nucleotide polymorphisms (SNPs)–variations at a single point in the genome–for just a few hundred dollars, he says.

Genetic samples were chosen to include individuals whose geographic ancestry could be determined, based on having all four grandparents coming from the same country.

The researchers then created a two-dimensional map with individuals positioned according to how similar or how different they are from all the others. When color-coded to show where each of their grandparents is from, the results are compelling, clearly showing the shape and boundaries of Europe.

One of the motivations for this kind of work is to assist genetic epidemiology, or population-wide genetic studies. Indeed, this is one of the main goals of Glaxo Smith-Kline, which participates in the study, says Novembre. “They are interested in pharmacogenetic purposes to do case control studies of adverse drug reactions,” he says.

“The idea is to save money in these large-scale genetic epidemiological studies,” says Krawczak. “It’s very costly to genotype people.” But if you can create genetic control groups for distinct populations, it allows you to more easily test drugs against different populations to see where the benefits lie, he says.

At the moment, the main focus is on Europe because it has a lot of genetic variation but a relatively well defined and delineated history. “It’s a nightmare to do population genetics in America,” says Krawczak. “There are so many migrant populations from different parts of the world that it’s just too complex.”

Even so, Novembre says that he plans to extend this sort of research to cover larger parts of the world and individuals of mixed ancestry. “At the moment, if you have mixed grandparent ancestry you appear between the set of countries where the grandparents come from,” he says. “So if they are part Italian and part British, they would appear in Switzerland. But we are working on algorithms that will be able to infer grandparent ancestry and get around this.”

Eventually, this sort of research is likely to be picked up by the growing number of companies offering DNA home tests over the Internet to people wishing to trace their genealogy. At the moment, these services tend to offer fairly rough pictures of one’s origins. But as the microarray technology becomes cheaper and the statistical software used to map it becomes more sophisticated, these kinds of services should greatly improve, Novembre says.