In the new study, researchers copied mutations from the 1918 and 1968 strains into the HA protein of the current H5N1 virus. They found that the same mutations that transformed the 1968 strain could also make the H5N1 HA protein better able to bind to human sugars. (Inducing the changes from the 1918 virus did not have this effect.)

Scientists were relieved to discover that the two mutations only partly transform the H5N1 protein -- it can bind to human receptors, but not as well as proteins from the 1918 or 1968 strains. "The ultimate conclusion, at least for mutations we know about, is that it's a lot harder to tailor H5 types of hemagglutinin to bind to human forms of receptors than it is for [other varieties] of influenza," says Jeffery Taubenberger, a scientist at the Armed Forces Institute of Pathology in Rockville, MD, who led the effort to sequence the 1918 virus.

Wilson adds that these changes have not been observed in the circulating H5N1 virus, and it's unclear what kind of impact they would have. "We're not saying it only takes those two changes to cross the barrier," he says. "You likely need a lot of other changes to get human virus."

Some experts are skeptical that the H5N1 strain will ever morph into a human virus. But they emphasize that, if it isn't H5N1, another deadly strain is likely to emerge, and better surveillance and analysis technologies are crucial for making the world better prepared for a pandemic flu, wherever it emerges. "As a screening tool, the [microarray] will become a really valuable test," says Taubenberger.

Government organizations are now considering using the technology to monitor the H5N1 virus as it spreads throughout the globe. So the technology could come in handy immediately.

The Scripps researchers are already using the technology to test new variants. An H5N1 strain isolated from people in Turkey was found to have some mutations in the HA receptor binding region. But, according to analysis with the glycan array, those changes did not boost the virus's ability to bind to human sugars.

Another new variant of H5N1, which began infecting people in Indonesia in 2005, was announced last week at the International Conference on Emerging Infectious Diseases in Atlanta, GA. According to scientists at the Centers for Disease Control, who led the research, the new variant does not appear to be more transmissible between humans, but it does have mutations in the crucial HA protein.

Says Anthony Fauci, director of the National Institute of Allergy and Infectious Disease in Bethesda, MD, "Understanding the molecular basis of the [virus's] ability to bind to various receptors gets us closer to understanding how viruses adapt to different species and has implications for understanding the response you want in a vaccine."