Posted for Declan Butler
An unpredecented aspect of the reaction of the scientific community to the current pandemic threat is the sheer speed with which researchers are making data publicly available. Within hours of the genomes of virus isolates having been analyzed, researchers from every corner of the globe have uploaded their sequences to the GISAID flu database, or Genbank, for anyone to compute.
Meanwhile, some journals have moved to warp speed, getting papers peer reviewed and published in days instead of months. Neil Ferguson’s group at Imperial College London, for example, published an initial report on the epidemiology of the outbreak in Science on 11 May (see ‘Swine flu spread matches previous flu pandemics’) . It used some sophisticated modelling to describe the evolution of the outbreak, even if the underlying epidemiological data available at that point to feed into the models was so scant that one leading public health blogger described the paper as “”https://scienceblogs.com/effectmeasure/2009/05/swine_flu_fast_track_publishin.php">computer-aided tea-leaf reading".
The New England Journal of Medicine (NEJM) also published a somewhat meatier paper on 7 May by scientists at the US Disease Control and Prevention (CDC) providing a useful summary of the clinical symptons and age distributions of the earliest cases (see ‘US swine flu cases dissected’).
But another group of leading evolutionary biologists, including Oliver Pybus at the University of Oxford, and Andrew Rambaut at the University of Edinburgh, have taken a completely different tack. While preparing papers for peer-reviewed publication, they have put online on a public Wiki sophisticated analyses of the flu genome, including detailed phylogenetic trees, as soon as they got their results. They argue that it is in the interest of the public and the scientific community to make data relevant to the pandemic threat publicly available as fast as possible.
And now, Science has just published a paper by another group covering much of the same ground. The paper has some 60 authors including scientists at CDC and from the World Health Organization’s lab network.
Like the open Wiki (which the paper doesn’t cite), the research retraces the genetic origins of the virus, showing that the virus originated from a mix of a North American and Eurasian strains of swine flu, which themselves contain avian and human flu genes. It includes some preliminary antigenic analyses of the virus, showing that they so far they seem homogeneous – that should simplify vaccine strain selection. It also shows that antibodies raised in ferrets to seasonal influenza
A(H1N1) did not cross-react with the new virus, so indicating that the existing seasonal flu vaccine would not protect people from the new virus.
The Science paper is “very useful” in that it collates all the available information, says Robert Webster, a flu virologist at St Jude Children’s Research Hospital in Memphis, Tennessee, adding however that he doesn’t feel it adds much new to what everyone in the field already has learnt from multiple sources over the past few weeks. Although he feels that it’s sad that the paper didn’t cite the public wiki, he points out that scientists are in an unusual and unfamiliar situation. It certainly raises issues of how best to align fast track open web publication with the peer review process during exceptional circumstances.
Derek Smith, a researcher who carried out the antigenicity part of the Science paper, says: “There is nothing special in the [phylogenetic] trees in the just published paper, some of the paper is getting the basic data out there, peer-reviewed.” As to his colleagues’ public wiki efforts, Smith has nothing but praise: ” I think it is really great that Oli, Eddie, Andrew, Gavin, et al have been posting [to the public wiki]. It’s also great that CDC have been publishing sequences, and much other information, on the day they are generated so everyone can see immediately what is happening, and do specialist analyses.”
Indeed much of the information now appearing i peer reviewed journals, or at least parts of it, has often been in the public domain beforehand, aired at the CDC and WHO’s daily press briefings, or published as Dispatches in the CDC’s Morbidity and Mortality Weekly Report. That latter venerable if abstract publication, usually read mainly by hardened public health nerds, has become required reading for anyone with a serious interest in following the latest developments of H1N1 swine flu.