Odyssey Program Poised to Revolutionize Research

To stay on the cutting edge of data analysis, Harvard researchers, from neurobiologists to climate scientists, are turning to Homer.

Assistant Dean for Research Computing James Cuff developed Odyssey, a centralized, high-computing system, when he arrived at Harvard in 2006, and scientists at Harvard and Massachusetts General Hospital are using the computing environment to process big data in their research, as computing sciences continue to figure prominently into medicine.

“In the style of one of the ancient Greek epic poems attributed to Homer, it was a journey!” Cuff wrote in an email. “The administration, senior faculty, and the deans each realized that without high quality research computing, Harvard wouldn't maintain their national and global competitiveness.”

The Research Computing cluster at Harvard, which was founded in 2007, offers computational resources for bioinformatic analysis, visualization, and data storage and currently serves 1215 active user accounts at three data centers.

Randy L. Buckner, a Harvard psychology and neuroscience professor and researcher at Massachusetts General Hospital, uses Odyssey to study human brain function in his lab.

“I think that one of the reasons why research computing and the Odyssey cluster is so important it that we’re just dealing with extraordinary amounts of data about individual brains, about large numbers of individuals, and even brain states over time,” Buckner said. “And to make sense of that complexity we need this level of computing horsepower.”

According to John D.E. Gabrieli, a professor of health sciences and technology and cognitive neuroscience at MIT, computational resources offer especially great insight into the brain and various neurological disorders, based on the brain’s inherent complexity.

“The Odyssey program at Harvard is invaluable in supporting the scientific needs of understanding the amazing computational powers of the brain, from neurons to large-scale networks,” Gabrieli wrote in an email.

While computer programs offer several advantages in processing big data, Buckner acknowledged concerns over security and data protection inherent in computing research as well.

“We’ve seen across the life sciences and medical fields in general appropriate concerns about privacy and medical records,” Buckner said. “Research computing that is working with human data, whether it be economics or medicine, is going to need to confront those challenges.”

—Staff writer Melanie Y. Fu can be reached at mfu@college.harvard.edu. Follow her on Twitter @MelanieYFu.