The visualization project did support previous lab work on transmission of the virus, however. Earlier studies on lab mice found that a certain genotype of the virus was particularly infectious. Looking at the mapped data, the researchers discovered that this genotype was quite prevalent in mammalian hosts--more so than they would expect if it occurred merely by chance. Janies says this suggests that this genotype is highly infectious in a variety of mammals--not just in mice.

While their initial research didn't provide a single host to target or an easy answer to combating the problem, Hill believes that the maps will raise new research questions and "inspire new hypotheses."

Janies hopes that more data on H5N1 will become publicly available so that the map can be improved. "There needs to be more public access to avian-flu data," he says. Scientists want to get credit for their work, so they hold back data until their papers are published, Janies says. And governments want to protect their economies. "If you're [a country that's] known to have avian flu," he says, "people won't travel there or buy your poultry."

The team is already applying its experience to other infectious diseases. A similar map for SARS is currently in development.

The team is already applying its experience to other infectious diseases. A similar map for SARS is currently in development.