Search

September 29, 2026

Setting Traps for Viruses

When Vaibhav Mohanty sat down for his Hertz Fellowship interview, he pitched an ambitious idea: What if he could use theoretical physics and chemistry to predict how a virus would evolve — and design around its next move before it made one?

Hertz bet on it. That bet has paid off.

A 2023 Hertz Fellow in the Harvard/MIT MD-PhD Program, Mohanty has developed a mathematical and computational framework for doing just that. Working on his second PhD in Professor Eugene Shakhnovich’s lab in Harvard’s department of chemistry and chemical biology, he is studying whether scientists can do more than anticipate viral evolution. His research asks whether they can reshape the evolutionary paths available to a virus, making dangerous mutations less likely to succeed in the first place.

The work won a national prize from the American Physical Society’s Division of Biological Physics in spring 2025 and was published summer 2026 in the Proceedings of the National Academy of Sciences.

Redesigning Evolution

For decades, evolutionary biologists have used fitness landscapes to map how mutations affect a virus’s ability to survive and spread.

Mohanty and Shakhnovich’s approach, fitness landscape design (FLD), explores whether those landscapes can be deliberately reshaped to make dangerous mutations less likely to succeed.

“Pathogens and cancers are constantly evolving to evade our immune systems,” Mohanty said. “FLD might help us fight back by proactively suppressing dangerous evolutionary trajectories before they appear.”

Using Physics to Stay Ahead of Viral Evolution

Think of evolution as a mountain range; viruses tend toward higher peaks, where they’re better able to survive and spread. Mohanty wants to know whether scientists can reshape the terrain — closing off routes that could lead to dangerous mutations.

In the vast “fitness landscape” of evolutionary possibilities, viruses try to mutate their proteins to improve their fitness and evade the human immune system. Researchers at Harvard and MIT have developed fitness landscape design (FLD) as a framework for designing optimal antibodies that can shut down these evolutionary paths, stopping viral evolution in its tracks before the mutations ever emerge. Source/artist: Vaibhav Mohanty (Harvard University and MIT)

Informed by his background in both physics and medicine, Mohanty developed two algorithms to do this: One searches for combinations of antibodies that could reshape the landscape; the other identifies a second antibody that could suppress escape mutations from the first.

Testing Fitness Landscape Design Against Real-World Data

To test the idea, Mohanty and Shakhnovich used simulations of viral evolution, norovirus experimental evolution data and SARS-CoV-2 chemical and epidemiological data to evaluate the FLD. The model closely tracked the evolutionary patterns.

They also applied FLD retrospectively to the COVID-19 pandemic. The analysis suggested that targeting a strain identified by the algorithm early in the pandemic could have constrained later escape variants while still protecting against the original strain.

The COVID-19 analysis is computational and depends on the model applying at population scale. Still, the analysis raises the possibility that proactive vaccines and antibodies could someday be designed with a virus’s future evolution in mind, not just the variants already circulating, setting the stage for new pandemic preparedness and biosecurity strategies.

Expanding FLD Beyond Viruses: AI, Cancer Therapies and Protein Design

FLD may also have uses beyond vaccines. Although these applications are still developing, Mohanty sees possibilities for the approach in cancer, where it could help researchers develop treatments that are harder for tumors to evade. The next generation of FLD algorithms may also help researchers engineer proteins faster.

For now, Mohanty is working on follow-up papers to understand which evolutionary paths can actually be shaped with a given set of antibodies and to incorporate the latest protein AI models into the existing physics-informed algorithms to accelerate FLD.

Full Circle: From Hertz Fellowship Pitch to Published Research

“It truly is so nice to see it come full circle after having this idea years ago, getting to talk about it in my Hertz Fellowship interview, and now having Hertz Foundation support me in seeing this work come to life,” he wrote to the Hertz Fellowship team.

About the Hertz Foundation

The Hertz Foundation is the nation’s preeminent nonprofit organization committed to advancing American scientific and technological leadership. For more than 60 years, it has stood as an unwavering pillar of independent support through the renowned Hertz Fellowship, cultivating a multidisciplinary network of innovators whose work has positively impacted millions of lives. Learn more at hertzfoundation.org.