|Dr. Mohammad El Abd and Dr. Jason Moore
(photo by Beth Hindmarsh)
By Beth Hindmarsh
What do computer games and the study of bee swarms have to do with cancer research? They are both pieces of the human disease jigsaw that, while seemingly unrelated, might yield powerful new tools of diagnosis and treatment. What they share is high-powered computing. Without it, computer games could not do the amazing things they do, solving how swarms of bees (or colonies of ants, schools of fish, or flocks of birds) seem to work as one entity would be impossible, and applying that knowledge to research into human disease processes, particularly cancer, would be unlikely to yield answers.
How high-powered computing might be applied to research in human disease was examined this summer by Mohammed El Abd, an assistant professor of computer engineering and expert in particle swarm optimization (PSO) at the American University of Kuwait, and Jason Moore, director of the Institute for Quantitative Biomedical Sciences (iQBS) at the Geisel School of Medicine at Dartmouth. El Abd is the fourth Faculty Fellow to study with a Dartmouth colleague. Previous Fellows include Simon O'Meara, assistant professor of art history; Amir Zeid, associate professor of computer science; and Mohamad Awad, assistant professor of electrical and computer engineering. The AUK-Dartmouth Faculty Fellow placement is awarded annually to an AUK researcher who identifies a Dartmouth colleague with whom to work.
According to Moore and El Abd, the development of increased speeds of graphical processing units (GPUs) for gaming has been instrumental in the management of the mountains of data that are generated each year. GPUs that are used in gaming contain many small processors that make them much more powerful than the central processing unit (CPU) of most computers.
By using algorithms (step-by-step procedures for calculations and data processing) to manipulate data and discover patterns that could not be seen with previous processing power, the potential for solving seemingly unsolvable problems is immense. What's amazing is that this has come about in only the last five or six years as graphical processing units have evolved, and they are upgraded just about every year.
El Abd and Moore met at the 2011 Genetic and Evolutionary Computation Conference (GECCO) in Dublin, Ireland. While they are coming from different perspectives, they have a common interest in high-performance computing. Says Moore, "I very much look forward to collaborating with Dr. El Abd on research projects at the interface of high-performance computing and evolutionary computation. I hope to visit AUK sometime in the next year to move these projects along."
"My visit to Dartmouth was highly successful," says El Abd. "I had access to a machine equipped with three GPUs, which enabled me to implement two versions of the PSO algorithm. Dr. Moore and I continue to talk about future collaboration."
Last Updated: 12/14/12