The World Health Organization has been warning for years that the world is overdue for a flu pandemic -- and that the avian flu virus circulating in Asia and Europe is the most likely culprit. Those concerns have grown with the impending arrival of bird flu in the United States this year. But, so far, the virus (H5N1), which is deadly to birds, has killed relatively few people.

In order to become a pandemic human virus, the H5N1 strain would need to mutate to become easily passed from person to person. Scientists don't know exactly what changes would allow the virus to make that deadly jump -- but they hope to find some clues with new technologies that can assess viral genomes and proteins with greater accuracy than ever before.

Viruses infiltrate host cells by binding to specific sugar molecules coating the cell. Human and bird cells carry different sugars, so viruses that infect birds don't usually infect humans.

However, viruses can evolve or swap genes -- many of the deadliest human flu viruses of the last century have had bird flu genes. The 1918 flu virus, for example, was a bird virus that acquired the ability to infect human cells and killed 30 to 50 million people worldwide. The viruses behind other more recent flu pandemics were human strains that acquired a few bird virus genes, making them more infectious to people.

Scientists hope that by studying the unique characteristics of previous pandemic flu strains, they will be able to predict how the current strain could mutate to become more deadly to humans. They have zeroed in on the hemagglutinin protein (HA), a molecule that sits on the outside of the influenza virus. The protein binds to sugars and determines which cells the virus can infiltrate.

In a Science paper published this month, researchers at The Scripps Research Institute in La Jolla, CA identified two mutations that could make the H5N1 virus more likely to infect human cells. "This a possible route by which the virus could...gain a foothold in the human population," says Ian Wilson, the Scripps scientist who led the research. "But it's not a doom or gloom story...I find it encouraging. Here's a possible avenue by which the virus can make the switch, and we should look out for it."

In previous research, the Scripps team tested how well the proteins from different viruses bind to human sugars, by using a new technology called a glycan microarray -- a glass slide coated with the different kinds of sugars found in bird and human cells. They discovered that two small mutations in the protein from the 1918 flu strain transformed that virus from one that binds only to bird cells to one that binds only to human cells. Two other small changes in the strain from the 1968 pandemic had a similar effect.