When it comes to responding to emergencies like pandemics and biothreats, time is of the essence.
An artificial intelligence project from the Defense Advanced Research Projects Agency, or DARPA, has the potential to help the world prepare for these types of crises.
Abhishek Singharoy, an associate professor in Arizona State University’s School of Molecular Sciences — currently on a special assignment with DARPA — is building a new AI program that can predict the function of every protein known to man in the span of about an hour.
The program, called NODES (Network of Optimal Dynamic Energy Signatures), combines physics with advanced artificial intelligence to analyze how proteins — the molecules that control how diseases spread and how our bodies respond — move and change shape.
While scientists know the sequence and structure of hundreds of millions of proteins, they only understand the function of a small fraction. By capturing and analyzing proteins in motion rather than relying on stationary protein structures, NODES can make high-confidence predictions about what the proteins do.
“If successful, NODES will dramatically improve our ability to assess potential threats from new and unknown protein sequences,” said Singharoy, who is also an associate faculty member of ASU’s Biodesign Center for Applied Structural Discovery. “It will also speed up the process of turning scientific discoveries into real-world applications, help us understand complex diseases, strengthen our defenses against infections and biothreats, and ultimately, protect human lives.”
Improving speed and accuracy
Current computational models rely on stationary protein structures, which limit accuracy and can only handle smaller datasets. NODES aims to change that by capturing how a protein moves and changes shape, and scaling up speed and optimization to evaluate roughly 300 million proteins in just one hour, with more than 90% confidence in its predictions.
We can think of current models as like trying to understand a movie by looking at a single still frame — we miss all the action and context.
NODES, on the other hand, watches the entire film at high speed, letting scientists see how proteins move and behave, and they do it for hundreds of millions of proteins in just an hour.
Why this research matters
Research is the invisible hand that powers America’s progress. It unlocks discoveries and creates opportunity. It develops new technologies and new ways of doing things.
Learn more about ASU discoveries that are contributing to changing the world and making America the world’s leading economic power at researchmatters.asu.edu.
“Abhishek’s work shows exactly the kind of cutting-edge interdisciplinary research we support at the School of Molecular Sciences,” said Tijana Rajh, professor and director of the school. “By pairing deep scientific expertise with tools like AI he and his team are pushing the boundaries of what we can discover and how we can protect human health.”
The speed and scope of NODES could prove critical in moments of crisis, from identifying potential bioweapons to quickly responding to a new infectious disease outbreak.
Imagine rapidly decoding the function of novel toxins before they can spread harm. The same technology could also reduce risks in biological applications, accelerate drug development and open doors to new treatments for complex illnesses.
“Having a program approved by DARPA is deeply meaningful to me because it reflects not just my own efforts, but also the support, collaboration and inspiration I’ve received from so many people along the way, including at ASU,” Singharoy said.
“It’s a reminder of the importance of innovation here at DARPA and motivates me to keep striving to make a positive impact. I’m grateful for the opportunity to continue contributing in the area of molecular biophysics.”