Faculty Focus: Network News
- Matt Spangler
Colin Campbell, assistant professor of physics, is the lead author of a study of network science published in the April 2017 issue of Scientific Reports.
Networks are everywhere. Think of neurons in the brain, signaling molecules inside cells in your body, disease transmission, or ecosystems. In each case, the system can be studied in terms of components (for instance, people who either have a disease or do not) and interactions (how and when people come into contact and possibly infect one another).
Whatever the system, says physics professor Colin Campbell, “there are some dynamics involved. The broad goal is to influence and control the behavior of the system. If we’re thinking about disease, we might say, ‘Let’s immunize people.’ If we’re talking about cancer, we might say, ‘How can I kill this tumor?’
“A network has many different components,” Campbell continues. “Now imagine an intervention that you can control. If you change one thing, for instance by introducing an abundance of a protein into a cell, you get a ripple effect in the behavior of the entire system. But it gets very complicated because if you have to change two or three things in sequence or in combination to achieve the desired outcome, all of the ripple effects overlap. There are these cascading influences. So, the theory that we’re working with is the theory of identifying how to route information through the system and what components you’re going to control directly in order to get the behavior that you want.”
In the paper, titled “Correlations in the degeneracy of structurally controllable topologies for networks,” Campbell and his colleagues, including his former thesis advisee Steve Aucott ’16, studied the properties of over 50 networks, representing diverse systems ranging from blood cancer to the structure of Internet.
They were able to find links between the pattern of interactions in a network and efficient methods for controlling its dynamic behavior. Campbell presented the work in 2016 at the Mathematical Biosciences Institute at Ohio State University, and it was published in Scientific Reports in April 2017.
“We were able to characterize the role of the components and the interactions in terms of information propagation,” Campbell explains. “So hopefully we’ve advanced the ball, so to speak, in terms of understanding the relationship between the network’s structure and dynamics and how they relate to controlling a system.”
The idea was developed in Aucott’s senior capstone project, titled “A high-dimension analysis of complex networks,” for which he won honors from the Department of Physics as well as the Department of Mathematics & Computer Science, where he worked with Shaun Ramsey, associate professor of computer science. Aucott was also the recipient of the John S. Toll Prize in Physics.
“I really like the applicability of network science,” says Aucott, a double major in physics and computer science who landed a job as a software engineer right out of college. “It seems totally abstract on first appearance, but it can be applied to a huge number of systems. I learned a lot from Professor Campbell because of his experience in a particular subset of complex networks.”
And Campbell’s research interest has inspired new coursework on scientific modeling and data analysis, which he first taught last spring.
“The class is all about using a computer for scientific ends; it’s an introduction to computer science for science students. I am able to talk about how we can use a computer to model these very complicated systems, because these systems are so big, so huge that there’s no way you can do these calculations by hand. When you have some network that has hundreds of components and thousands of interactions, the computer becomes a necessary tool.”
As a first-year graduate student of physics at Penn State, Colin Campbell didn’t know much about biology or computer programming, but when he sat in on a series of faculty research talks, one topic—on complex biological systems—quickly piqued his interest. “I thought I would do condensed matter physics, because that’s what I was most familiar with as an undergrad. But I became fascinated by networks and how they can be used to model biological systems. The idea of joining an interdisciplinary team to apply new methods to the study of biological systems was really fascinating.”