U.S. scientists have discovered a hidden molecular “switch” that herpes viruses rely on to invade cells. By combining AI, simulations, and lab experiments, the researchers identified and altered a single amino acid that shut down viral entry.
According to Washington State University virologists, what once might have taken years was achieved far faster using computational tools. These findings open new possibilities for designing future antiviral treatments.
The study focused on uncovering and blocking a specific molecular interaction that herpes viruses rely on to gain access to cells. The process of invading cells by a virus is very complex, and there many interactions. Not every interaction is equally important — most of them are a form of ‘biological background noise- – but there are some critical interactions, which the AI has helped to identify.
Viral Fusion Process
The researchers examined a viral “fusion” protein that herpes viruses use to merge with and enter cells, a process responsible for many infections. Science continues to have only a limited insight into how this large and complex protein changes shape to make cell entry possible, which explains why vaccines for these widespread viruses have been difficult to develop.
To tackle this challenge, the researchers turned to artificial intelligence and detailed molecular simulations. Professors Prashanta Dutta and Jin Liu analyzed thousands of potential interactions within the protein to identify a single amino acid that plays an essential role in viral entry.
The two academics created an algorithm to examine interactions among amino acids, the basic components of proteins, and then applied machine learning to sort through them and pinpoint the most influential ones.
AI computes success
After identifying a key amino acid, the research team moved to laboratory experiments. By introducing a targeted mutation to this amino acid, it was found that the virus could no longer successfully fuse with cells. As a result, the herpes virus was blocked from entering the cells altogether.
According to Liu, the use of simulations and machine learning was essential because experimentally testing even a single interaction can take months. Narrowing down the most important interaction ahead of time made the experimental work far more efficient.
“It was just a single interaction from thousands of interactions. If we don’t do the simulation and instead did this work by trial and error, it could have taken years to find,” Liu explains. “The combination of theoretical computational work with the experiments is so efficient and can accelerate the discovery of these important biological interactions.”
More to learn
Although the scientists confirmed the importance of this specific interaction, many questions remain about how the mutation changes the structure of the full fusion protein. The researchers plan to continue using simulations and machine learning to better understand how small molecular changes ripple through the entire protein.
“There is a gap between what the experimentalists see and what we can see in the simulation,” Liu clarifies. “The next step is how this small interaction affects the structural change at larger scales. That is also very challenging for us.”
The research appears in the journal Nanoscale, titled “Modulation of specific interactions within a viral fusion protein predicted from machine learning blocks membrane fusion.”
