Grant funds computer scientist's look into the hidden architecture of DNA

A computer scientist is combining physics, biology, and artificial intelligence (AI) to tackle some of medicine’s most complex challenges, from designing better drugs to understanding how aging may begin deep inside human cells.

Alexey Onufriev, professor in the Department of Computer Science, has received a National Institutes of Health (NIH) Outstanding Investigator Award (MIRA), worth about $2 million over five years, to advance computational modeling research with implications for cancer, age-related disease, and drug discovery.

“Computer science is increasingly most powerful at the boundaries where our tools and methods meet problems we didn't originate,” said Christine Julien, head of computer science. “This NIH MIRA Award is a signal from the biomedical research community that Alexey's work belongs at the center of their most important questions."

Flexible funding, nimble science

Unlike traditional project-specific grants, this award funds a broader scientific vision, allowing investigators flexibility to pursue new discoveries as research evolves.

“This is an award that funds a lab, not a specific project,” Onufriev said. “It allows us to follow the science where it leads.”

At the center of the research is a challenge fundamental to nearly every area of biology and medicine: understanding how molecules move and interact inside living systems. Onufriev plans to use the grant to support research into improving the accuracy and speed of molecular simulations, AI-assisted drug design, and the emerging science of how DNA organization inside cells may influence biological function and aging. 

Proteins, DNA, water molecules, and drug compounds are constantly shifting, binding, and changing shape inside cells. Onufriev’s lab develops computational methods that simulate those molecular interactions accurately and efficiently enough to help other scientists better understand disease and predict how potential treatments might behave before they ever reach a laboratory.

“How do you design a molecule that binds well?” Onufriev said. “You need to be able to model the motion and the interaction of molecules.”

One major focus of the lab is using these molecular simulations for computer-aided drug discovery.

Traditionally, pharmaceutical researchers experimentally test thousands of compounds to identify molecules capable of blocking disease-related proteins. The process is costly, time-consuming, and often unsuccessful.

Onufriev’s team is developing computational techniques that can screen millions of possible molecules virtually before researchers move into laboratory testing. This allows identification of stronger drug candidates earlier in the process, accelerating research into treatments for a range of diseases.

The lab is also exploring how artificial intelligence can improve molecular modeling while remaining grounded in the laws of physics.

“What we’re trying to do differently is combine physics with machine learning,” Onufriev said. “You have the physics, which guarantees that you don’t get complete nonsense, and you have machine learning to get the details right.”

Desciphering the hidden genetic code

Another major component of the grant project focuses on what Onufriev calls a “second genetic code.” This is the idea that biology is shaped not only by DNA itself, but by how DNA is physically packaged inside cells.

Each human cell contains roughly six feet of DNA compressed into a microscopic nucleus. Scientists increasingly believe the organization of that DNA determines which genes become accessible and active, influencing everything from cell specialization to disease development.

“The DNA in the nose and the tongue is exactly the same,” Onufriev said. “How it folds itself determines how it functions. So somewhere else, additional information is hidden.”

That hidden layer of information may also help explain aging.

Emerging research suggests aging may not be driven solely by accumulated DNA damage, but by disruptions in how DNA is organized and accessed inside cells over time. Onufriev’s group is developing computational models to better understand those structural changes and explore whether restoring more youthful DNA organization could reverse some effects of aging.

The lab's team is currently studying those questions using computational models and fruit fly systems in collaboration with researchers in entomology.

The research reflects the increasingly interdisciplinary nature of modern computer science. Onufriev’s lab brings together collaborators and students from computer science, physics, biology, and related fields to solve problems that span multiple scientific disciplines.

Graduate, undergraduate, and even high school students are already contributing to projects connected to cancer biology, DNA accessibility, and molecular simulations.

Pursuing science for the public good

For Onufriev, the award provides something increasingly rare in academic research: long-term stability to pursue ambitious scientific questions.

While the research could eventually contribute to breakthroughs in medicine, drug development, and aging science, Onufriev said he intends for the work to remain patent-free and openly available to the scientific community.

“If you patent something, then nobody can work on it. Then, what’s the point?” Onufriev said. “You have an idea, you put it out there, others build on it. It’s a very collaborative thing.”

Using the lab’s computational tools and discoveries, the intent is to accelerate progress across many areas of biology and medicine.

“Ultimately, science is about spreading knowledge,” he said.