Computer Modeling May Unlock Answers to Complex Problems like Heart Disease
June 14, 2011
What does the beating of your heart have to do with a computer?
Ask Alan Caceres.
As a child, Caceres was curious about what made electronic gadgets actually work. Now, as a computer science major at Lehman College, he is uncovering their secrets—which led him to take a high-powered workshop earlier this year with Professor Nancy Griffeth of Lehman's computer science faculty.
Although the title—Workshop on Atrial Fibrillation—would suggest an audience of cardiologists, those attending actually comprised fifteen students of varying majors who came from many senior colleges within the City University of New York.
The workshop was the second in a series designed to introduce students to the work underway by Professor Griffeth and her colleagues at several prestigious research institutions, including Carnegie Mellon, Cornell, and NYU. They are developing models of biological and man-made systems, as well as techniques for building and analyzing these models, that could help us better understand certain challenging problems.
One of those problems is atrial fibrillation—an abnormal heart rhythm. Other challenges being studied include pancreatic cancer, automobile safety, and aerospace safety. By engaging students like Caceres in this work, Professor Griffeth hopes to encourage them to continue their own scientific research. The project is funded by the National Science Foundation.
At the workshop, Caceres was introduced to the very computationally intensive research being down with NVIDIA graphics cards. In the consumer world, the many parallel processors on these cards are used to improve picture quality on everything from downloaded movies to 3D video games, but NVIDIA provides the capability to use the processors for computation as well as for controlling the display. As a result, the cards can be used to do the time-consuming computations needed to study complex biological systems, like the heart.
In addition, Caceres, who is entering his senior year at Lehman, learned about various simple oscillators—which alternate between different states for long periods of time, similar to a beating heart—and how they tie into the specific research underway. After completing the workshop, he had "great respect" for the other disciplines represented in the sessions and also understood the need for more powerful computing. He was inspired to continue research in his field because "there are obstacles that can be overcome in the near future that can help further expand any interdisciplinary research topic."
Caceres, who is receiving the Lehman College Computer Science and Mathematics Scholarship, is part of the LSAMP (Louis Stokes Alliance for Minority Participation) fellowship and received a second-place overall trophy for mathematics at a recent LSAMP conference. He is interning at IBM, has already produced two research publications, and plans to pursue a doctorate in human-computer-interaction (HCI) and then possibly a doctorate in computational physics.
During the previous workshop in 2010, which focused on pancreatic cancer, nutrition major Tamara Schillin used both math and the BioNetGen tool suite to map signaling pathways in cells, which are involved in the development of cancer. She participated in the workshop to contribute her knowledge of biology and to learn about biochemistry—she was planning to take a course in that area during the following semester.
"The students worked in teams to map the pathways," Schillin explained, "and each student had strengths in math, science, computers, or all of those fields. It was an extremely rigorous and satisfying experience, and one that I won't soon forget."
Schillin sees a direct connection between this kind of work and a significant public-health outcome. "Understanding the myriad of reactions and the kinetics allows scientists to make predictions about products and outcomes," she notes, "and knowing which mechanisms to slow down, switch-off, or accelerate is key to curing diseases like cancer and diabetes."
Schillin sees computers as "the best tool for aiding scientists in navigating the many complexities and variances of cellular metabolism," and that applies to nutritionists as well, who seek to understand how various nutrients and molecules are metabolized and their role in various diseases. The technological modeling produced in the workshop, she believes, "would be ideal for clinical nutritionists, to assist them in making diagnoses and recommendations." The experience, she says, gave her a glimpse into the future.
The subject of atrial fibrillation was inspired by the work of Flavio Fenton of Cornell University, who spent a week teaching and working with the students at the workshop, and Scott Smolka and Radu Grosu of Stony Brook University. The pancreatic cancer workshop was based on the work of Edmund Clarke and Christopher Langmead at Carnegie Mellon University and James Faeder of the University of Pittsburgh. The programs used in that workshop were developed and explained to the students by Ezio Bartocci, a post-doctoral student at Stony Brook University.