Computer network aids crucial drug research

By Kate Shepherd

The University of Chicago has joined forces with IBM, Argonne Labs, and the University of Texas Medical Branch (UTMB) to develop drugs that will treat and cure deadly diseases such as dengue, West Nile virus, and Hepatitis C.

The project will let researchers tap into massive amounts of computer processing power made available by computer owners around the world who will volunteer their idle computers’ processing capabilities. By making use of the increased processing capacity, the project will test potential drug compounds for their viability as treatments to stop the diseases, typically a time-consuming process.

“IBM has the hardware and systems software expertise. UTMB has expertise in protein structure. We at U of Chicago and Argonne have the expertise in binding free energy calculation and grid computing,” said Yuqing Deng, a computational scientist at Argonne. “It is wonderful to see the research in all these different fields—computer science, computational chemistry, and structural biology—work together to make a real impact in drug discovery.”

The project is currently in a preliminary phase, but the collaborating institutions have secured computers and grid-computing software, according to Deng. He says researchers are working to develop an accurate and efficient way to predict the bonding affinity of drug-like compounds to target proteins.

The most effective drugs are ones that bond with the most specificity, said Benoit Roux, co-principal investigator of the University’s Institute for Molecular Pediatric Sciences. Roux said that computing the bonding affinity of a drug is necessary to determine its effectiveness. While such computations can be time-consuming, researchers working on this project hope to speed up the process while maintaining the quality of their work.

“Typically, this process is worked on by trial and error in the pharmaceutical industry, but we are working on it computationally,” said Roux. “It’s a kind of adventurous way to approach it, but we hope it is a good approach. It has not been tried before because people do not believe it will work.”

Interested individuals can download software that will allow the World Community Grid, a virtual supercomputer to run computations on users’ idle computers from the Grid website. Running on a single computer, it would likely take about 50,000 years of computational time to finish the needed calculations, but if the project can pool together the idle computer time it requires from across the world, the research could be completed in less than a year. The Search for Extra-Terrestrial Intelligence (SETI) project has utilized a similar system, with over three million users donating idle computer time to the project.

“Without World Community Grid, we would have to make inexact, simplifying assumptions that have proven to be obstacles to previous drug development efforts,” said Deng. “World Community Grid enables us to perform comprehensive calculations that yield accurate biochemical results and therefore give us the best chance to discover cures for these serious worldwide diseases.”