The Department of Biochemistry welcomes Hannah Wayment-Steele, who joined the faculty on August 18, 2025. The Wayment-Steele Lab will build deep learning models (a type of machine learning) that better predict the dynamics and functions of biomolecules, and will then test these models with experiments.
Wayment-Steele majored in math and chemistry as an undergraduate student at Pomona College in California. At the time, the concepts of biochemistry seemed a world away from her research interests.
While completing her master’s degree in chemistry at Cambridge University in the United Kingdom, Wayment-Steele came to appreciate the connection between these realms of science, and the chemical complexities of biomolecules. “The thing that caught my interest was that, in order to function as they do biologically, proteins and RNA have to occupy different conformations and change among those conformations with specific timings. It’s like a dance,” Wayment-Steele says. “A single sequence can have enough information to dictate all of the moves in this dance.”
With a newfound appreciation for biomolecules, Wayment-Steele shifted her focus. During her doctoral research at Stanford University, she modeled the dynamics of proteins using computer simulations and developed more effective statistical methods to understand the models. Her research also used experimental data on thousands of RNA sequences to build a model that could better predict RNA ensembles, work that proved useful in the race to find an effective, shelf-stable mRNA vaccine for SARS-CoV-2.
Wayment-Steele explains that RNA is an intrinsically unstable molecule, which makes it difficult to develop mRNA vaccines with a capacity to be stored for long periods. Using computational methods, Wayment-Steele and other researchers identified different mRNA sequences that can fold in ways that improve RNA stability.
In her postdoctoral research at Brandeis University, Wayment-Steele created a large dataset that held information she and her colleagues used to train a model predicting protein dynamics. They were interested in information that could help identify new protein targets for treating diseases, obtained by predicting the locations of protein dynamics that were too slow for existing computer simulations to reach.
Now, as an assistant professor at UW–Madison, Wayment-Steele and her lab will continue to build and improve on deep learning tools available to scientists. The first steps, she says, are to creatively identify what questions biochemists would like to be able to answer, then to identify existing data sets that might hold information to answer that question.
“Deep learning is great once you’ve identified the task you want to optimize. But there are so many interesting questions [in biochemistry] that haven’t even been formulated for deep learning yet. That’s what interests me the most — it’s the step before data can be used to build a new model,” says Wayment-Steele.
Wayment-Steele sees the breadth of biochemical wisdom and experience at UW–Madison, as well as the emphasis on entrepreneurship and translational research, as a key to her lab’s success. “I want our research to have real-world impacts. Through collaborations on campus, I’m excited to apply our models beyond what I alone could ever think to do,” she says.
Wayment-Steele has already begun to grow her lab, and a postdoctoral researcher will be joining the lab this year. An avid rower, Wayment-Steele sees her role in the lab as akin to a coach. “When we’re tackling a research question, I don’t know what the answers are from the start,” Wayment-Steele says. “My role is to watch the game’s progress, to give insights and pointers, to be there when people need me. And, to make sure they know that they can work hard and be a world-class scientist and they can have interests and passions outside of their work hours.”
Written by Renata Solan.