New DNA map will speed search for disease genes

NEW YORK -- Scientists have mapped patterns of tiny DNA differences that distinguish one person from another, a step that will speed up the search for genes that promote common illnesses such as heart disease and diabetes.

The map represents "a real sea change in how we study the genetics of disease," said Dr. David Altshuler, a leader of the project that included more than 200 researchers from six nations.

Scientists want to find disease-related genes as a means for diagnosis, prediction and developing treatments.

Genes that predispose people to common disorders -- heart disease, cancer, diabetes, asthma and others -- are devilishly hard to find.

But the new "HapMap" -- taken from the genetics term haplotype -- opens the door to launching comprehensive searches through the human DNA for those genes, said Altshuler.

It's not yet clear how many genes will be found or how useful they will be, Altshuler said. The achievement probably won't result in new disease treatments for five or 10 years or more, he said.

HapMap data already have been used to track down a gene linked to the eye disorder macular degeneration, said HapMap project participant Dr. Lincoln Stein of the Cold Spring Harbor Laboratory in New York.

The map is described in Thursday's issue of the journal Nature. Participating scientists came from Canada, China, Japan, Nigeria, the United Kingdom and the United States. The project analyzed DNA samples from 269 people from Nigeria, Beijing, Tokyo and Utah.

Human DNA consists of more than 3 billion building blocks whose sequences form genes, just like letters spell a word. For any two unrelated people, these letters are 99.9 percent the same. But that leaves millions of single-letter differences, called SNPs (pronounced "snips") that provide genetic variation between people.

The HapMap shows how more than 1 million of these SNPs form patterns: A person with one particular version of a SNP is highly likely to carry particular versions of other SNPs as well. (A group of variants that associate together is called a haplotype.)

This clustering greatly simplifies the task of analyzing what variations a person carries, because not all of them have to be identified.

And this, in turn, allows scientists to use great arrays of SNPs as signposts for tracing genes that promote disease or determine a person's response to a particular medicine.

It's not clear yet how the project will pay off for human health, David Goldstein and Gianpiero Cavalleri of Duke University wrote in a Nature editorial. But it should greatly help scientists find genetic variants that predict a patient's response to particular treatments, which can one day be used to guide patient care, they said.