The first published map of human genetic differences offers a major step toward truly personalized medicine, from predicting who will get what disease to finding ways of choosing the best drug for an individual patient, scientists said on Thursday.

The map, published on Thursday, looks at more than 1.5 million tiny genetic differences among 71 people, said the team at Perlegen Sciences, Inc., a spinoff of Mountain View, California-based gene chip maker Affymetrix.

This is enough to find some of the most common genetic variations involved in disease, said David Cox, Perlegen’s chief scientific officer.

“This project sets a new milestone in the search for genetic elements linked to complex genetic diseases such as Alzheimer’s, cancer and multiple sclerosis,” he said.

The idea is not to find one or two “disease genes” but instead to look for patterns that can account for how a person may respond to specific high blood pressure drugs, for instance, Cox and colleagues report in Friday’s issue of the journal Science.

“It’s kind of like being a matchmaker,” Cox told a news conference at the annual meeting of the American Association for the Advancement of Science. Eventually, he hopes, individual doctors will be able to consult such a map to determine which drug to prescribe to a patient.

This will build on existing policy at the Food and Drug Administration, said the agency’s Dr. Lawrence Lesko, an expert in pharmacogenomics - the field of using genetic patterns to predict drug reactions.

Already, targeted cancer drugs are approved for a narrow group of patients, such as Breast Cancer patients eligible to use Herceptin, he said. “You are excluding using the drug in a non-responder,” Lesko told reporters.

PREVENTING TOXIC RESPONSES

A map like Perlegen’s might provide the first step toward screening out those in a population likely to have a toxic response to a drug, he said.

One example is the COX-2 inhibitors, analgesics being reviewed this week by FDA experts because they raise the risk of heart attacks, strokes and heart disease in some patients.

To make the map, Perlegen worked with researchers at the California Institute for Telecommunications and Information Technology (Calit2) at the University of California San Diego, and the University of California at Berkeley’s International Computer Science Institute.

They scanned 71 Americans of African, European and Asian descent, picking out 1.58 million of the most common single-letter variations in the genetic code, called single-nucleotide polymorphisms.

The genetic map is made up of combinations of nucleotides known as A, C, T and G. Different combinations of these spell out the amino acids that in turn make up proteins used in the cells and by the body.

The current maps of the human genome, published in 2001, use combinations of human genes, all combined together to make a picture of the “average” human map.

But to understand why one person has high blood pressure, or why another cannot tolerate aspirin, the differences must be teased out.

This map is a very small start, Cox said.

In a commentary, David Altshuler of Harvard University and the Massachusetts Institute of Technology warned that not all drug reactions and disease outcomes will boil down to genetics. Environment is sure to play a role, he said.

“The predictive value (of genes) may not be high,” he told the news conference.