Fruit Flies, Ants and Algorithms
For years, Saket Navlakha (Ph.D. ’10, computer science) has been fascinated by biological processes in nature. But when he looks at a plant or an ant or even a fruit fly, he sees much more than biology. As a computer scientist, he also sees the potential to discover new algorithms.
“All of biology, I view it as one big computer,” explained Navlakha, an associate professor at Cold Spring Harbor Laboratory in New York. “It’s not like the computers that we use every day, but it’s fundamentally computation and problem-solving. And everything I see is sort of tilted by that lens now.”
In his research, Navlakha explores what he calls “algorithms in nature,” how collections of molecules, cells, and organisms process information and solve computational problems that are crucial for survival. He views biological systems—in everything from growing plants to tiny insects to the disease-fighting systems in our bodies—as problem-solving machines.
“When you look at the problems that biology has to solve, there are a lot of parallels between those problems and the problems humans face,” Navlakha said. “We think some of those biological solutions could be useful in practical applications, since biology is so distributed, energy efficient and adaptive.”
In dozens of published studies over the past several years, Navlakha identified algorithms in nature that can advance computer science. It’s an intriguing research perspective that won him two major honors in the past five years: the National Science Foundation Faculty Early Career Development Program (CAREER) Award and the Pew Biomedical Scholars Award.
“The ability to have my feet firmly in both these very different directions tells me this interface we’re building is something that people in the computer science world and the biology world are finding cool,” Navlakha said. “We’re speaking a language that resonates with both communities.”
Computer science was part of Navlakha’s DNA from the very beginning. His father, a professor and former director of the School of Computer Science at Florida International University, steered him in that direction early on. But it wasn’t exactly love at first sight.
“We got a computer when I was in elementary school, so I was exposed to the internet and programming very early,” Navlakha recalled. “But I wasn’t in love with computer science back then and it wasn’t till much later that I became really passionate about it.”
Navlakha majored in computer science at Cornell University and worked his way through the required courses. But he felt uninspired until a class on information retrieval sparked his interest.
“It basically told us how to apply theory to build something like Google, this massive, really cool search system that can really efficiently find relevant documents,” he said. “That piqued my interest, and I started getting more into it after that.”
Soon after graduating from Cornell with a bachelor’s degree in computer science in 2004 and a master’s degree in computer science in 2005, Navlakha co-founded Gabbly, a startup aimed at finding a way for people visiting the same website at the same time to communicate with each other. After six months, Navlakha realized he missed being in an academic environment and reached out to universities about Ph.D. opportunities.
Finding his niche
In fall 2006, when Navlakha started his Ph.D. in computer science at the University of Maryland, he knew he wanted to study networks. But he didn’t find his niche until he met Carl Kingsford, a computer science assistant professor at the time with a research focus in computational biology. Kingsford’s class on graphs and networks gave Navlakha his first introduction to biological systems and an exciting direction for his Ph.D. research: the study of protein interaction networks.
“These are like social networks, but instead of people, you have proteins,” Navlakha explained. “There are thousands of these proteins in our cells forming complex interaction networks that are involved in basically every biological process, and we were interested in answering some basic questions about the structure and evolution of these networks.”
After earning his Ph.D. in 2010, Navlakha moved on to Carnegie Mellon University as a postdoc. It was there that he began to look deeper into biological systems and their problem-solving mechanisms. And he wondered what computer science could learn from biology. It’s a question he continued to explore in his research at Carnegie Mellon, then as an assistant professor at the Salk Institute for Biological Studies in San Diego after that.
“It gave me this purpose, I could really be a voice to plants and insects and understand them from a more computational and algorithmic perspective,” Navlakha said. “It broadened my interest in biology because I think all of biology is computing.”
By 2019, Navlakha had returned to the East Coast, settled down with his wife, a physician at Memorial Sloan Kettering in New York, and accepted a position as an associate professor at Cold Spring Harbor Laboratory, where his research continues to unravel algorithms in nature and their potential applications.
“One of our studies is looking at neuroscience and neural circuits as inspiration for new machine learning algorithms. In another study, we’re looking at ways plants form 3D branching structures and using them as inspiration for building geometric networks for various engineering problems,” Navlakha explained. “Recently we’ve been thinking about the immune system, basically the firewall that detects and keeps away pathogens and viruses, and we’re trying to understand the algorithms that make that happen.”
Plants, insects and algorithmic solutions
By studying plants, insects and more, Navlakha learned that sometimes the smallest biological systems can yield big answers to challenging problems in computer science. For example, he discovered that fruit flies can teach us a thing or two about search engines and how to make them more efficient. He found that these tiny insects have a unique system that allows them to sort good food odors from bad ones by comparing them to smells they’ve encountered before.
“We studied the neural circuits in the fruit fly brain and discovered a new algorithm for doing this type of similarity search,“ Navlakha explained. “A lot of people took the basic algorithm we published and improved on it, and today they’re using these algorithms for similarity searches for medical images and things like that.”
For Navlakha, this work at the intersection of biology and computer science is part of a much bigger mission, one he hopes to pursue for many years to come.
“I want to publish an encyclopedia of algorithms that support life,” he explained. “I think I would die happy if I could do that, identify all these natural algorithms that have evolved in the immune system, in plants, in the brain, and other systems. It’s much more than any single lab can do, but I hope to at least contribute to it.”
Written by Leslie Miller
The Department welcomes comments, suggestions and corrections. Send email to editor [-at-] cs [dot] umd [dot] edu.