Public health experts have warned that controversial experiments on mutant viruses could endanger lives by unleashing an accidental pandemic.
Several groups of scientists around the world are creating and altering viruses to understand how natural strains might evolve into more lethal forms that spread easily among humans.
But in a report published on Tuesday, researchers at Harvard and Yale universities in the US argue that the benefits of the work are outweighed by the risk of pathogenic strains escaping from laboratories and spreading.
They calculate that if 10 high-containment labs in the US performed such experiments for 10 years, the chance of at least one person becoming infected was nearly 20 per cent. If an infected person left the laboratory, the virus might spread more widely.
"We are not saying this is going to happen, but when the potential is a pandemic, even a small chance is something you have to weigh very heavily," said Marc Lipsitch, an epidemiologist at Harvard School of Public Health, who wrote the report with Alison Galvani, an epidemiologist at Yale.
The report threatens to reignite a crisis in science that erupted in 2012 when a US biosecurity panel ruled that two separate studies on mutant bird flu were too dangerous to publish. They described the creation of new mutant strains. One fear was that the recipe for the pathogens might fall into the hands of terrorists seeking biological weapons.
Those studies, led by Ron Fouchier at Erasmus medical centre in Rotterdam, and Yoshihiro Kawaoka at the University of Wisconsin-Madison respectively, were eventually published after months of delays. Other researchers have now begun similar experiments.
Both Fouchier and Kawaoka criticised the latest report, published in Plos Medicine, and said their work had full ethical, safety and security approval, with risks and benefits accounted for.
Last year, the US government, which funds most of the controversial work, revised its guidelines for so-called dual-use research of concern.
Under the new rules, work can be funded if the potential benefits from it are substantial and the risks considered to be manageable.
But Lipsitch said there was no evidence that the risks and benefits had been weighed properly. "To my knowledge, no such thing has been done, but funding for these experiments continues," he said.
Simon Wain-Hobson, a virologist at the Pasteur Institute in Paris, said scientists working on the controversial virus studies should be less defensive.
"There are times when we have to open up and face our critics. Marc is articulating what many of us feel is obvious," he said.