News

An undergraduate student and his professor at Illinois Institute of Technology (Illinois Tech) recently used Expanse at the San Diego Supercomputer Center, part of the UC San Diego School of Computing, Information and Data Sciences (SCIDS), to develop patent-pending technology that could speed up drug research.

David Cooper, an Illinois Tech undergraduate, and his professor, Robert E. Frey Jr. Endowed Chair in Chemistry David Minh, invented an algorithm—using Expanse and two other U.S. National Science Foundation (NSF) ACCESS program supercomputers—that helps to predict which cellular pathways a drug will activate, and how much it will activate them. The algorithm could accelerate the design of drugs that are safer and more effective by avoiding pathways that could cause side effects.

The team has published their work in the Journal of Chemical Information and Modeling and cofounded a company, Biagon Inc., to develop and apply the technology to industrial drug discovery projects. Cooper has since completed his college degree and now works full-time as the CEO of Biagon.

The project is being continued as one of the first to receive funding under a newly formed NSF-funded Accelerated Research Translation (ART) award at Illinois Tech. The ART award helps researchers at institutes of higher education to significantly elevate the level of research translation for economic and societal impact, by providing funding for “seed translational research projects.” The term “translation” refers to the process of “translating” research from the lab into the consumer marketplace.

“The Accelerating Research Translation program…identifies, and champions institutions positioned to expand their research translation capacity by investing in activities essential to move results to practice,” NSF Director Sethuraman Panchanathan said in a news release, adding, “We were pleasantly surprised to find that one of our first funded seed translational research projects would result in the formation of a startup that is also led by an undergraduate.” 

Biagon’s machine learning algorithm, whose patent by Cooper and Minh is pending, will potentially help pharmaceutical and biotech companies develop safer drugs faster—with fewer side effects and shorter research runways. In essence, the algorithm can identify safer, effective drug compounds more quickly than current drug research processes allow.

“The use of graphical processing unit computing resources from SDSC Expanse was critical to performing the work described in our first publication,” said Minh. “We would not be able to perform adequate molecular dynamics simulations without access to such resources.”

In addition to Expanse at SDSC, the team utilized ACCESS allocations on Bridges-2 at the Pittsburgh Supercomputing Center and Jetstream2 at Indiana University. Computation resources are supported by NSF ACCESS (award no. BIO240146).