Terabytes, exabytes, yottabytes: bytes by the zillion are heading this way, thanks to the appetites of particle physicists, Earth scientists and organismal and molecular biologists (see Briefing on pages 517-520). So how will researchers cope?

The particle physicists are well versed in handling such data. They seem to be crossing their fingers and hoping that the off-the-shelf hardware and software will be available in time to allow them to analyse the copious output of their next-generation colliders. Given the history of the technology, their expectations are well founded.

Molecular biologists, on the other hand, appear to have eyes for data that are bigger than their stomachs. As genomes near completion, as DNA arrays on chips begin to reveal patterns of gene sequences and expression, as researchers embark on characterizing all known proteins, the anticipated flood of data vastly exceeds in scale anything biologists have been used to.

Biologists are waking up to the challenge, albeit belatedly. Last week, for example, an expert panel told Harold Varmus, director of the US National Institutes of Health (NIH), that his agency needs to do much more in this direction. As panel co-chair David Botstein rightly emphasized, the big issue is training. Nevertheless, sheer supercomputer power (and, for that matter, medium-sized computer power, too) will be essential. To that end, the panel recommended support for the development of existing supercomputing facilities established for other disciplines, as opposed to new facilities for biology. The strain on existing facilities is enormous, so serious investment will be required, but, the panel argues, not in a way that re-invents wheels. This approach also has the advantage of leveraging NIH dollars at centres that already house needed expertise and infrastructure.

But, equally urgently, the NIH needs to help develop a new generation of computer-wise researchers. The panel recommended that five to twenty “National Programs of Excellence in Biomedical Computing” be established. It's a measure of the speed of the revolution, and of the dearth of expertise in the community, that it's not obvious where in the United States such programmes would be launched. Just as urgent is the need for computer specialists in laboratories, and a change in attitude that sees them as invaluable rather than second-class citizens. In short, computer experts at $85,000 a year are, like it or not, an increasingly necessary component of grant applications.

It would be wrong to leave the US agenda in this area solely in the hands of the NIH. The National Science Foundation has traditionally been the focus of support for supercomputing in the research community, while the Department of Energy has supported most of the country's particle physicists. Both agencies have made their own investment in informatics, and sharing what they learn with biologists is as urgent a necessity as increasing NIH spending.

Other countries and regions face similar shortages of skills. Meanwhile, private companies are busily sequencing and computing and licensing. There is, therefore, an urgent underlying message that all scientifically ambitious countries should heed: strong government funding for the quantitative analysis of data is essential if the results of fundamental biological research are to remain a public good.