Arieh Warshel has studied enzymatic reactions his entire career. Credit: Chris Shinn

For decades, Arieh Warshel, USC Distinguished Professor of Chemistry and a 2013 Nobel laureate, has used computer simulations to understand how enzymes—fundamental to nearly every biological process in living organisms—carry out the essential work of life.

But about five years ago, Warshel’s simulations began to hit their limits. The simulations weren’t accurate enough to answer the questions he cared about most—especially how mutations change the behavior of enzymes, which enable viruses to replicate and spread. So, he shifted course, determined to solve this puzzle.

"I never leave a problem," he said. "I just attack it from different directions."

This time, that direction was artificial intelligence.

Warshel’s first AI-based effort focused on trying to predict how mutations in enzymes alter their activity. In the case of viruses like HIV, which causes AIDS, or HCV, the virus that causes hepatitis C, mutations enable viruses to "escape" drug treatment and continue replicating.

His group discovered a striking pattern: Across several studies, the speed of an enzyme’s activity correlated strongly with a statistical measure called "maximum entropy." The breakthrough meant they could use a purely statistical and computational approach to determine the maximum entropy—thus, predicting enzyme function.

The finding opened a door. If maximum entropy could predict enzyme behavior, perhaps it could also predict how viruses outmaneuver drugs.

Warshel had wrestled with this challenge before. In 2008, he built computer models to forecast which mutations might help HIV escape treatment. The work had some success, but it couldn’t anticipate the virus’s next move, and it required enormous computational effort to test possibilities one at a time. The question lingered: Was there a better way?

An inventive childhood

Warshel, born in 1940, traces his grit and determination back to growing up in Israel on a kibbutz where his mother worked as a beekeeper and his father pioneered the development of fishponds, an important part of the kibbutz economy. In his book "From Kibbutz Fishponds to The Nobel Prize, Warshel" describes a free-range childhood where he devised and tested parachutes for cats and built a small balsa wood airplane powered by actual flies.

Later, in the Israeli army, he became skilled in Morse code. Always with an eye on university, he lugged physics textbooks with him as he lived out of a tent and took part in brigade maneuvers. He was accepted into Technion, the Israel Institute of Technology, and decided to major in chemistry. By his third year, he was studying enzymatic reactions, the field Warshel would pursue for the rest of his life.

Applying maximum entropy

Recently, he tried applying maximum entropy to the problem of how viruses evolve to evade drugs. HIV seemed like the natural place to start, not only because it was the focus of his early career, but because Stanford University maintains a vast database of thousands of HIV mutations that appear in patients who are taking different drugs.

The results, he said, were interesting, but ultimately discouraging. Maximum entropy did correlate with patterns of drug resistance. Unfortunately, so did a much simpler measure: the raw number of mutations the virus accumulated. "HIV protease mutates in almost any way you can imagine," Warshel said. "Because of that, it’s incredibly difficult to predict what it will do next." At best, the approach could compare drugs already tested in patients, but clinicians could usually see those patterns firsthand.

So, the team expanded their search. Instead of the most mutationally explosive virus known, they turned to pathogens with more constrained evolutionary landscapes. One of the first was hepatitis C virus (HCV). There, the picture changed. Maximum entropy aligned much more cleanly with the mutations the virus actually adopted under drug pressure. That opened the possibility of forecasting its "next move," as Warshel put it—a way of playing chess with the virus, using both the strength of each mutation and its likelihood.

The results of this study were recently published in the Proceedings of the National Academy of Sciences.

Drug resistance isn’t the only frontier. As the team applied maximum entropy beyond viruses, they found it worked remarkably well in other biological problems—including diseases linked to myosin, a molecular motor essential for muscle contraction, heart function and hearing.

Mutations in different myosin proteins can cause forms of hereditary deafness or abnormalities in the heart muscle. In these systems, maximum entropy showed strong, reproducible correlations with disease-causing mutations and with the effects of drugs acting on the proteins. Compared to the brute-force simulations Warshel had relied on for decades, the method was easier, faster and in many cases more accurate.

"These are much lower-hanging fruits," he said. "It’s working very well."

A new paper from his group, now under review, reports that maximum entropy can even predict how effectively myosin "walks"—a description of how it generates movement along cellular tracks when altered by specific mutations. It’s the latest in a series of publications showing how the method can illuminate complex biological behavior with unexpected precision.

Drug resistance remains an important challenge for Warshel, even if HIV itself is too unruly a system to conquer with current tools. But the broader lesson, he said, is that maximum entropy has become a powerful lens for understanding how biological systems respond to mutation—one that may ultimately reshape research across fields.

"We keep pushing on drug resistance," Warshel said. "But we’re having much more success in enzyme design and in predicting diseases like heart conditions and hearing loss. Maximum entropy works beautifully for these problems."

More information: Linfeng Hu et al, Exploring evolutionary trajectories of drug resistance, Proceedings of the National Academy of Sciences (2025). DOI: 10.1073/pnas.2517715122

Citation: Decoding the chemistry of life: Maximum entropy reveals how mutations alter enzymes and drive drug resistance (2025, December 9) retrieved 9 December 2025 from https://phys.org/news/2025-12-decoding-chemistry-life-maximum-entropy.html

This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no part may be reproduced without the written permission. The content is provided for information purposes only.