Bridging Biology and Computer Science to Understand Cellular Differentiation
Manu Setty, Computational Biologist
We all begin life as a single cell. A cell that must repeatedly grow and divide to create all of the cells that make up our body. As these cells divide, they must reshape themselves into the many different types required to build a human, such as a heart cell or a brain cell. This process, called cellular differentiation, is how we go from a single cell to a complex organism with distinct organs and tissues.
Cellular differentiation is not only crucial for embryonic development, but it is required throughout our lives to maintain the tissues that make up our bodies: As old cells die off, new cells need to replace them. Disruptions to the complex chemical signaling pathways that regulate cellular differentiation can lead to abnormal development, blood disorders, and cancer. The desire to understand how and why cells differentiate — and what can go awry in disease — is the driving force behind Dr. Manu Setty’s research.
Even as a young student, Setty recalled having two primary academic interests: biology and computer programming.
“I have always been driven by a curiosity of how things work,” he said.
As Setty looked toward college, he was confronted with a decision: biology or programming. He noted that at the time, studying biology was expected to lead to a career in medicine. Less passionate about becoming a physician, Setty ultimately decided to pursue computer science at the National Institute of Technology in Surathkal, India. After receiving his bachelor’s, he spent a couple of years working as a programmer but eventually discovered it wasn’t as fulfilling as he hoped.
A desire to combine his long-held passion for biology with his programming skills led him to New York City, where he pursued a master’s degree in computational biology at Columbia University. Computational biology sits at the intersection of computer science and biology, allowing researchers to use computational tools to answer biological questions.
“I’m a strong believer in using technology as a driver for discovery. My work has been about embracing new types of data because it can give us new types of information, expanding the questions we can ask biologically,” Setty said.
During his master’s program, he had the opportunity to collaborate with Dr. Christina Leslie at the nearby Sloan Kettering Institute, the research arm of Memorial Sloan Kettering Cancer Center. While working with Leslie, he first experienced what would become his research passion, understanding how gene regulation drives a cell to change from one type to another.
“I really felt that was my calling. I found it super interesting,” said Setty, who would go on to join Leslie’s lab for his doctoral work.
Throughout his research career and post-doctoral work, he continued to focus on cellular differentiation, working to predict why cells would change into one type and not another and potentially determine which changes may lead to disease.
“My main focus was developing algorithms using single-cell data to understand developmental trajectories and cellular differentiation,” he said.
When looking to start his own lab, Setty was drawn to Fred Hutch because of the exciting work being conducted and the potential for collaboration with colleagues.
“I fell in love with the research environment,” he said. “It’s very foundational, very creative.”
“An understanding of the healthy system functions as an anchor for understanding disease. If you have a reference for how things should work in a healthy controlled environment, it helps us pinpoint where, how, and particularly why things are going wrong when it comes to disease.”
Setty is a member of both the Basic Sciences Division and the Translational Data Science IRC. He likes the freedom this gives him to pursue collaborations across disciplines. His lab is housed in Fred Hutch’s state-of-the-art Steam Plant facility, where he uses advanced genomics techniques and computational algorithms to make groundbreaking insights into cellular differentiation.
“How a cell makes its decision is incredibly fundamental, but I feel the question is still incredibly enigmatic,” he said.
In order to answer these challenging questions, Setty and his team use machine learning, a set of algorithms that can improve themselves through experience. This approach can be particularly useful when dealing with large amounts of noisy data with complex interactions.
“Machine learning can help you identify rules and patterns which you cannot identify by hand, and in many cases, you can’t even write a formula for because the relationships between these variables are super complex,” he said.
Despite the power of machine learning, Setty says it is essential to have a firm grasp of the underlying biology.
“While you have machine learning in your toolkit, you really need to understand the biological systems you’re studying and model your data in the right way to answer questions,” he said.
The answers gleaned from his work provide both a deeper understanding of fundamental processes of our biology and insights into blood cell disorders and tumor formation. In diseases like cancer, once-specialized cells can regress and regain characteristics of their earlier selves, possibly leading to excessive growth.
Setty says that in order to understand what goes wrong in disease, it’s crucial to know how the system should function normally.
“An understanding of the healthy system functions as an anchor for understanding disease. If you have a reference for how things should work in a healthy controlled environment, it helps us pinpoint where, how, and particularly why things are going wrong when it comes to disease,” he said.
— By Matthew Ross, October 15, 2021
RELATED RESOURCES
