The lab-engineered H5N1 bird flu virus whose recipe the U.S. government doesn’t want published may be less lethal than originally reported. But that fact, revealed at a scientific conference in Washington in the past week, isn’t likely to change many people’s minds about whether details about the bug should be kept secret or made public.

What is indisputable is that the virus, a close relative of one that has killed about 60 percent of the people it has infected in the past 15 years, is able to pass easily between mammals, not just birds. Even if it turns out to be only a fraction as virulent as its wild relative, that’s enough to make it necessary keep information about how to make it under wraps, many experts think.

“This has always been about transmissibility,” said Michael T. Osterholm, an epidemiologist at the University of Minnesota’s Center for Infectious Disease Research and Policy. “Have we added a new species to which this virus can be readily transmitted? The evidence suggests we have.”

“The big question is how does the transmissibility of the engineered virus compare to the starting [wild] one,” said David A. Relman, a physician and microbiologist at Stanford University.

Osterholm and Relman are on the National Science Advisory Board for Biosecurity, a 23-member committee that advises the federal government on issues involving biological research that can be used for nefarious purposes. In December, the panel asked the journals Science and Nature to hold up publication of two papers about experiments that made the bird flu virus contagious in ferrets, the lab animals used to mimic human influenza infections.

The journals complied, with the understanding they would eventually publish the research either in redacted form or entirely, with editorial additions. The nearly unprecedented case of self-censorship has since been the hottest topic in science on both sides of the Atlantic.

Both journals have published commentaries for and against publication. The World Health Organization held a two-day meeting in Geneva last month to discuss the issue. The Royal Society, in London, is holding a two-day conference in April; the National Academies, in Washington, a one-day meeting in May.

The biosecurity panel decided this week to meet again, this time with the lead authors of the two papers present to query them further about their findings. That will probably happen later this month.

“It wouldn’t surprise me if at least the majority, and possibly everyone, still saw this work the same way it did in December,” Relman said.

The description of the engineered virus’s lethality provided in the past week is either a clarified version of what appears in one of the unpublished manuscripts or a revised version. Opinions differ. It’s impossible to know which view is right because the only people who know the details won’t or can’t talk about them.

The gist of the initial report was that two research teams — one in the Netherlands, the other in Wisconsin, and both funded by the U.S. government — had taken “wild-type” H5N1 and modified it so that it could pass through the air from an infected ferret to an uninfected one in an abutting cage. This was reportedly achieved without altering the bug’s capacity to kill. Globally, H5N1’s lethality for humans is 59 percent — 345 deaths out of 584 confirmed cases. Most of the victims have been poultry workers in Southeast Asia; a few have been family members who cared for the victims.

At a meeting of the American Society for Microbiology on Wednesday, the leader of the Dutch team, Ron A.M. Fouchier of Erasmus Medical Center in Rotterdam, provided more particulars about his work.

He said the team infected six ferrets with wild virus and six with engineered virus by spraying the bug directly down their throats into their windpipes. All of the animals in both groups died. The team then sprayed an unusually high dose of each virus into the nasal passages of ferrets. Two animals that got the wild virus by this route died, but the engineered virus was fatal to only one in eight animals. The team then put ferrets infected with the engineered virus next to caged, uninfected animals. Seven animals became infected, but none died (the number of exposed animals in this scenario was not clear from the presentation).

“So this virus, although it is highly lethal to chickens and highly lethal if you put it directly down into the respiratory tract at high titer [concentration], it is certainly not highly lethal if ferrets start coughing and sneezing to one another,” Fouchier told the gathering.

The presentation raised two questions.

First: Did the fact that none of the ferrets infected by the airborne route die exonerate the virus of the charge of being highly lethal as well as obviously transmissible?

Not necessarily, said numerous people (only a few of whom would speak on the record). Both the wild and engineered strains were highly lethal when sprayed into the windpipe, which is evidence that the engineered version wasn’t materially weakened, they argue. Further, the number of animals in all the experiments was small, making it impossible to draw confident conclusions about the different outcomes in the other experimental conditions.

“We don’t know in the natural world how it gets into humans,” Relman said of wild H5N1 bird flu. “We don’t know whether it requires direct contact or large droplets or whatever. But we do know that the wild virus can kill humans.”

“Even if this virus is 20 times less virulent than the wild type, you still have a pandemic worse than 1918,” Osterholm said. The “Spanish influenza” killed about 2 percent of its victims in 1918 and 1919, a total of at least 50 million people worldwide.

The second question, less important, is: Was this information in the manuscript Fouchier submitted to Science?

At the meeting Wednesday, Anthony S. Fauci, head of the National Institute of Allergy and Infectious Diseases, said he had heard “old data that was clarified and new data that was juxtaposed with old data” at that meeting and also at the one in Geneva, where Fouchier also spoke.

Not everyone agrees.

“This is not a big surprise to me,” Relman said of the revelation that none of the airborne-infected animals died. “I can easily see that what he now clarifies as his procedure could be understood from what he wrote in the paper.”

He added, “But the paper was not clear.”

Fouchier could not be reached for comment by either phone or e-mail. Yoshihiro Kawaoka, the head of the Wisconsin team, has previously said his transmissible virus was not lethal in ferrets.

Both scientists have agreed to submit rewritten manuscripts. They will reportedly include, among other things, more discussion of the risks and benefits of this type of research.