PICSciE conference explores cross-disciplinary collaborations that accelerate scientific discovery

Thursday, May 30, 2019
by Melissa B. Moss

The Princeton Institute for Computational Science & Engineering (PICSciE) held its sixth biennial conference on May 10 in Lewis Science Library, with this year’s theme being an exploration of cross-disciplinary collaborations that accelerate scientific discovery. The all-day event was organized by PICSciE and co-sponsored by the Center for Statistics & Machine Learning, the Office of the Dean for Research,  and OIT Research Computing. Featured speakers included Peter Elmer, Physics; Barbara Engelhardt, Computer Science; Shantenu Jha, Computer Engineering, Rutgers; Sebastian Seung, Princeton Neuroscience Institute; Amit Singer, Program in Applied and Computational Mathematics; and Mengdi Wang, Operations Research and Financial Engineering.

Cross-disciplinary collaborations are happening more than ever, said Sebastian Seung, Evnin Professor in Neuroscience, professor of computer science and the Princeton Neuroscience Institute, whose conference talk was titled “Petascale Neural Circuit Reconstruction.” Seung combines methods from biology and computer science to extract the structure of neural circuits from light and electron microscopic images, and then applies computational methods to explain how they function.

“Neuroscience has entered the era of big data,” said Seung. “A mere cubic millimeter of brain tissue can yield a petabyte of electron microscopic images. To handle such complex and voluminous datasets, we make use of deep learning and distributed computing.”

“The goal of the conferences is to bring the broader Princeton research computing community together for a day of interesting talks, posters, and conversations,” explained Jeroen Tromp, Director of PICSciE and Blair Professor of Geology, professor of geosciences and applied and computational mathematics. “This year we attempted to highlight collaborative computational research, recognizing the fact that much modern research requires close collaborations between domain scientists and computer/computational scientists.”

Said Amit Singer, professor of mathematics and a member of the executive committee of the Program in Applied and Computational Mathematics, collaborations also serve to strengthen the intellectual connections between researchers with different backgrounds and the graduate and postdoctoral students they mentor.

PICSciE Spring 2019 conference

Prof. Barbara Engelhardt presenting a talk titled "Low dimensional representations of high-dimensional genomic data: Finding patterns that lead to discovery"

“I did my doctoral studies in applied mathematics at Tel Aviv University during which I spent about six months as a visitor of the Department of Molecular Biophysics and Physiology at Rush University Medical Center in Chicago,” noted Singer. “There, I was mentored by Bob Eisenberg, the chair of the department at the time and a leading researcher in the field of ion channels. I learned a lot from the collaboration with Bob, not only from his expertise in biophysics and the importance of mathematical modeling, but more generally how to communicate with biologists and researchers outside my field. Our collaboration influenced the way I choose problems to work on and how to manage my research group. Bob was extremely enthusiastic about science in general and I was lucky to be in an environment that encouraged scientific independence. These are things that I take with me ever since.”

Barbara Engelhardt, associate professor of computer science, opened the conference with her talk, “Low Dimensional Representations of High Dimensional Genomic Data: Finding Patterns that Lead to Discovery.” Engelhardt reflected on how collaborative opportunities come about at Princeton: “I think researchers at Princeton benefit enormously from both the smallness of the campus and the extraordinary intersectional centers and organizations here. The smallness of the campus means that I interact with people from every department regularly, and collaborations form very naturally. The intersectional organizations like PICSciE both catalyze these interactions and enable this work to be performed in a seamless way.”

“Over 100 people registered for the conference, representing 27 academic departments/programs at Princeton, as well as attendees from Columbia, Johns Hopkins, NYU and Rutgers. It was a unique opportunity for world-class computational researchers to share the impact of their research to the public,” said event organizer Ma. Florevel Fusin-Wischusen, institute manager of PICSciE. “It’s fascinating to hear how cross-disciplinary collaborations have truly accelerated scientific discoveries in recent years, and Princeton is at the forefront of it.”