Experiments with a deadly flu virus, suspended last year after a fierce global debate over safety, will start up again in some laboratories, probably within the next few weeks.

The research touched off a firestorm in 2011 when it became known that two groups, one in the Netherlands and another in the United States, had genetically altered a dangerous bird flu virus to make it more contagious in mammals.

Some scientists warned that a deadly pandemic could break out if the mutant virus leaked out of the lab accidentally or if terrorists stole it or made it themselves, using articles in scientific journals for the recipe.

The outcry led scientists conducting the experiments to declare a voluntary moratorium, in part to let research organizations and governments decide what safety rules to require.

Now, researchers say, the moratorium should end because most countries have rules in place. A letter from 40 scientists -- the same ones who called the moratorium last year -- was published Wednesday in the journals Science and Nature, saying it is time for the work to begin again in countries ready to allow it.

But the United States, which pays for much of the flu research both at home and abroad, has not yet released new guidelines. So U.S. scientists will not be able to resume experiments yet nor will those in other countries who depend on U.S. grant money.

Dr. Anthony Fauci, director of the National Institute of Allergy and Infectious Diseases, said that the Department of Health and Human Services was reviewing new guidelines and that he expected them to be approved in weeks. The guidelines will specify the laboratory conditions under which this type of research is permitted and require that experiments have a potential benefit for public health.

The experiments involve a type of bird flu virus called H5N1. It does not often infect people but appears unusually deadly when it does.

So far, H5N1 has rarely spread from person to person. People who fall ill have nearly always caught it from poultry. But flu viruses mutate a lot, and the fear has been that H5N1 will somehow become more contagious in humans.

The debated experiments involved ferrets, which are considered a good model for flu research because they react to the virus in much the way people do. Ferrets normally do not infect one another. However, by genetically manipulating the virus, researchers created a form that became airborne and spread from ferret to ferret. Its transmissibility set off alarms.

Advocates of the research say it is necessary so scientists can recognize changes in naturally occurring viruses that are dangerous and signal the need to eradicate infected animal populations.