The US journal Science published research Thursday on how a mutant bird flu may spread among mammals and possibly humans, following months of controversy over the risks of bioterrorism.

The paper detailed how a Dutch lab engineered an H5N1 bird flu virus that can be transmitted in the air among ferrets, and followed the publication last month of findings by a US-based team that made similar advances.

Last year, a US biosecurity panel called for only heavily edited results of the two papers to be released, for fear that an ill-intentioned scientist might be able to use the data to unleash a potent and lethal form of bird flu that humans could catch easily.

But international experts have since agreed that the benefits of publishing outweighed the risks.

Deadly flu pandemics have killed millions of people in the past. Until now, there have been fewer than 600 human cases of H5N1 bird flu infection in the world since it first infected people in Hong Kong in 1997, but more than half of all cases have been fatal.

The World Health Organization has tallied 606 human cases of bird flu since 2003 and 357 deaths, according to its latest report issued this month.

Lead researcher Ron Fouchier, a scientist at the Erasmus Medical Center in the Netherlands, said the aim was to gain a better understanding of how avian flu is spread in order to prepare for a potential human outbreak.

"The virus did not kill the ferrets that were infected via the aerosol route," said Fouchier, who has frequently stressed that the dangers of his research were overblown in the media.

"Anybody with access to the scientific literature can read all about dangerous pathogens that are more interesting to terrorize the world with than our particular virus."

Instead, Fouchier and his colleagues showed that the H5N1 virus could become airborne among ferrets -- considered a reasonable but not perfect model for humans -- after as few as five mutations and without mixing H5N1 with another flu virus.

The previous paper by Yoshihiro Kawaoka at the University of Wisconsin and colleagues, published in May in the British journal Nature, described how the virus could become airborne after a series of mutations and re-assortments with the 2009 H1N1 virus, or "swine flu."

The two papers offer important insights into what forms a spreadable bird flu may take, and could lead to more advances in how to stop it, said Anthony Fauci, director of the National Institute of Allergy and Infectious Diseases.

"I believe that the benefits are greater than the risks. Does that mean there's no risk? No, of course not," he said.

"Being in the free and open literature would make it much easier to get a lot of the good guys involved than the risk of getting the rare bad guy involved."

But it remains unclear just how much risk people face, according to Derek Smith from the University of Cambridge, who co-authored an article in the same edition of Science on the potential for such a virus to evolve in humans.

Smith said his team's research determined that viruses with two of the mutations are already being found in birds.

To reach the minimum level of five mutations that Fouchier's team described looks "pretty difficult, but we don't yet know how difficult it is," he said.

"We now know we are living on a fault line. What we have discovered in this working collaboration with Drs Fouchier and Kawaoka is that it is an active fault line," he told reporters.

Asked if such a virus would inevitably evolve in nature, Smith likened that to asking, "Could it ever snow in the Sahara?"

"It is absolutely within the realm of possibility that they could evolve in a human host or some other mammalian host. We see nothing -- we see no fundamental hurdle to that happening," he added.

In the meantime, Fauci said a voluntary moratorium on research that involves the causes or transmissibility of H5N1 has yet to be lifted, as leading US health officials try to establish rules for future experiments that could raise alarm among biosecurity experts.

"We are still struggling a bit," said Fauci.

"I can't tell you when it's going to be voluntarily lifted, but we are working very hard right now (to establish a) broad general criteria of the kinds of experiments that could be done."

The editor-in-chief of Science, Bruce Alberts, said the eight-month-long controversy "has shone a spotlight on the need to deal more effectively with 'dual-use research of concern.'"