• 3/14/2007
  • web-based article
  • staff
  • ComputerWorld (www.computerworld.com)

A project that allows the public to donate idle computer time to a grid computing system is helping scientists more quickly solve complex medical problems. For example, the World Community Grid has helped researchers dramatically speed efforts to develop new drugs to treat HIV and to identify new treatment paths for people with several different types of cancer.

IBM launched the community grid in late 2004 to allow computer users worldwide to donate idle computer processing power to a grid tasked with performing medical research. Since then, its users have donated the equivalent of more than 78,000 years of research time, according to IBM.

The grid today includes 265,000 members and 530,000 devices. It has helped researchers compile about 60 million research equations, officials said.

The Help Defeat Cancer project has used the grid to help build a massive public database of tissue samples that eventually will be used like a national fingerprint registry to get more accurate diagnoses from biopsies.

David Foran, professor of pathology and radiology and director of the Center for Biomedical Imaging at the University of Medicine and Dentistry of New Jersey (UMDNJ), said the effort has already created the “gold standard of databases” for breast, head and neck, and colon cancer. The project aims to eventually use the database to diagnose cancer and to plan the best course of treatment for patients, said Foran, who is leading the project.

In addition to the tissue images, the database stores comprehensive data about treatments received by patients participating in the project. As a result, physicians and pathologists can compare a new biopsy to all those in the database and then determine which types of drugs and treatments have worked best for other patients with the same types of tumors, Foran said.

It would have been impossible for Foran’s research team — which has no access to a supercomputer — to create the database without the community grid, he added.

“Going forward we could probably create a similar gold standard database for any number of areas of cancer research,” he said. “It literally would take years in computation for us to accomplish what we did in the first month on the grid, and we are only using a fraction of the grid.”

The project is sponsored by the Cancer Institute of New Jersey, UMDNJ Robert Wood Johnson Medical School, Rutgers University and the University of Pennsylvania.

The Fight AIDS @ Home project, sponsored by The Scripps Research Institute in La Jolla, Calif., is using the community grid to help speed its research efforts. The goal of this project is to use the computational power of the grid to find out which chemicals can best block the advance of HIV. Those chemicals could then be used in medicines to treat AIDS, officials said.

The research project is very compute-intensive, explained Arthur Olson, professor in the department of molecular biology at Scripps, because researchers have to test the effectiveness of certain chemicals against the virus itself and against the multiple mutations of the virus.

“If a virus is present and you treat it with a drug, the mutations that will help it survive in the presence of that drug are the ones that will dominate,” he said. “That is the continuing problem in AIDS therapies now — patients do develop drug resistance.”

For the first phase of the AIDS project, which launched in November 2005, researchers performed the equivalent of five years of research in six months because of the grid, Olson said. Scripps already has identified 40 chemicals that merit further testing, he said.

For the second phase of the project, which began in April and is ongoing, researchers are using the grid to study how to prevent the virus from forming in the body.

The first phase alone would have taken 10 years or more to run on Scripps’ supercomputer, Olson said. He noted that the experiments the researchers are running would not have been possible without the grid.

Next on the agenda of Olson’s group is using the grid to try to come up with a drug to be used against the variants of avian flu that are evolving.