The international pharmaceutical industry has been phenomenally successful over the past couple of decades in decreasing the burden of human suffering - at least for those of us fortunate enough to live in the prosperous market economies of the developed countries. As one might expect in a market economy, this success has been reflected commercially, in growing revenues and strong share prices. However, the international pharmaceutical companies are facing a profound problem: there are too few new products in their pipelines to sustain the industry at its present size into the future.
The companies have been aware of this for some time. The pattern of mergers and acquisitions that has resulted, for example, in Glaxo SmithKline is a consequence of this inexorable commercial logic: if there are too many companies and too few new compounds, then decrease the number of companies. However, this path has limited attraction. Why not try the other route of increasing the number of new compounds coming through clinical trials and on to the market?
A decade ago, the human genome project was being hailed as the saviour of the industry. Genomics, it was believed, would uncover all manner of new proteins that could serve as 'targets' for intervention with small molecule drugs. It has not worked out like that. By increasing the numbers of targets, genomics has encouraged greater spending at the research end of the drug discovery chain. But the commercial problem lies with validating targets, and then with checking that the small-molecule chosen to lock on to the target does not have toxic side-effects that are discovered only in the final stages of clinical trials, once vast sums of money have been spent.
Scientific computing has a significant role to play in helping solve these problems. At one level, Laboratory Information Management Systems (LIMS) and their extension into clinical trials management software can cut costs. As the story on our news pages describes, providing patients with electronic diaries can facilitate data collection in clinical trials. At another level, as the feature article by Peter Rees shows, information technology applied to the drug discovery process can help pharmaceutical companies manage the mountains of information being generated by genomics and by high throughput experiments. But in contrast to the euphoria that greeted the genome project, the reception accorded to computing in drug discovery has been muted.
Perhaps everyone is a bit wiser nowadays and no longer treats each new technology as some panacea that will cure all (commercial) ills. But there is another reason, one that might appear somewhat strange to those who are specialists in scientific computing: it's really rather difficult. There are structural difficulties for large companies to integrate all their data across all their sites, so that results are available to all. But there are also difficulties on the individual level. Biologists and chemists are not always at ease with computing. Simultaneously, they expect too much of the technology and they do not know what to ask of it.
There is an immense task here, not just for the pharma companies themselves, but for the whole of the scientific computing community. The direct payoff will be commercial, but indirectly it will contribute to the noble end of diminishing the burden of disease.
Dr Tom Wilkie
