Sir

In publishing the Letter to Nature by D. Kobasa and colleagues, “Enhanced virulence of influenza A viruses with the haemagglutinin of the 1918 pandemic virus” (Nature 431, 703–707; 2004), Nature has endorsed a publication that reports the creation of an influenza strain with increased virulence (at least in mice) based on the molecular structures of one of the deadliest diseases of the twentieth century. This will surely bring both benefits and risks to biosecurity.

Following a discussion of the problems arising from the publication of biosecurity-sensitive data, scientific journal editors came together and agreed that this issue deserves attention and that some general measures should be in place (see Nature 421, 771; 2003). Since then, most leading journals, including Nature, have introduced procedures to deal with this issue.

However, there are good reasons to believe that these individual journal-specific procedures are inadequate, in that they give the least restrictive journal the ultimate control over sensitive biosecurity information. Looking at the tremendous impact the outcome of the editor's decision may have on the public, informed democratic participation in the decision-making process must be a requirement. At present, none of these journals release information on the risk–benefit analysis undertaken in specific cases to allow independent assessments.

As long as clear guidance from legislators is missing, the policies followed by these scientific journals will remain non-transparent, possibly inconsistent and subject to political bias. Let's hope that the initiatives being pursued by several countries, including the United States (http://www.aaas.org/spp/post911/agents), to negotiate biosecurity guidelines for scientists, will lead to workable and publicly accepted principles regarding all aspects of research, including the publication of research results.

A necessary first step towards transparency could be the publication of the local biosafety committee's reasons for giving approval, the biosafety measures taken and the editor's risk–benefit assessments.