Researchers at Mount Allison University
are working to better understand how human DNA is damaged and what drugs can be used to repair it, using a powerful IBM
Linux supercomputing cluster. The research will reduce the time it takes to develop drug treatments for diseases such as cancer and Alzheimer’s.
The 84-node, dual processor cluster of IBM eServer xSeries running Linux, allows researchers at Mount Allison to easily study how environmental factors such as UV light and cigarette smoke damage human DNA. Using this data, the researchers then study how human molecules ‘virtually react’ when exposed to various drugs.
“Because most of our computational chemistry research takes place during the summer when there are no classes, we needed a powerful system that could produce an enormous amount of data in a relatively short period of time,” says Stacey Wetmore, a chemistry professor and the project’s lead researcher. “The more complex the molecules being studied, the longer these calculations take, but our cluster produces data as fast as we can use it.”
Researchers can use the new cluster to process numerous reactions daily, considerably more and with greater accuracy than was possible with the 4-processor system used previously. Studying these types of drug reactions in a virtual environment will significantly speed drug discovery and is more cost effective than traditional laboratory experiments as it eliminates the need to purchase and pay for the disposal of chemicals.
“IBM’s expertise in Linux cluster and in computational chemistry applications helped us design and implement this server cluster,” says Dominic Lam, National High Performance Computing Manager at IBM Canada. “With our help, Mount Allison has been able to conduct their research far more accurately and effectively than in the past. They’re now doing things that just weren’t possible before.”
The Mount Allison Cluster for Advanced Research was funded by the Canada Foundation for Innovation and New Brunswick Innovation Foundation. In the future, it will also be used to build control applications for large, distributed IT systems and for high-energy physics research.