Medicine and the Internet Computing for Cancer: the role of distributed computing in cancer research

Computers have always been used in medicine for patient management, personal research and communication. Recently, the utilization of massively parallel processing, brought about by distributed computing grids, has allowed for cancer researchers to develop drugs and test therapies at speeds never before available. The use of distributed computing in testing molecules for their cancer protein binding potential, automated tissue microarray analysis and tumour growth modeling are all currently promising research areas. Despite some inherent challenges, distributed computing may one day be used extensively in all areas of medical research. This article has been reviewed by Dr. Eric Wong and by Dr. Wayne Weston and by Dr. Victor Han. The Internet may be synonymous with browsing and email, but its utility is much greater. The Internet is often used to garner information, be it a patient searching for facts regarding a recent diagnosis, or a physician trying to find the latest published research on treatment for an obscure disease. A conceptual link which is not immediately made, however, is the applicability of computers and the Internet in producing potential breakthroughs in cancer research. Cancer research has traditionally focused on exploring epidemiology, practical experiments in molecular biology, clinical trials of potential medications or therapies and, more recently, an exploration of immunotherapy and gene therapy. It is a direct extension of cancer research’s foray into molecular biology and genetics which has ushered in the use of computers in the oncological arena. History of distributed computing Since the birth of computers over fifty years ago, the raw speed of computation has increased by a factor of over one million. 1 Despite this increase in computational power, personal computers are still considered extremely slow in terms of requirements for solving highly complex scientific problems. 1 Cancer research is one such area where the complexity of molecular and biological interactions practically precludes the use of even modern personal computers in any but the simplest experiments. One method of solving this problem emerged in the 1980’s, and involved clustering multiple individual computers to form what was termed