(Nanowerk News) Identifying whether and how nanoparticles and proteins will bind to each other is a critical step in being able to design antibiotics and antivirals on demand, and a computer model developed at the University of Michigan can do that.
This new tool could help find ways to stop antibiotic-resistant infections and new viruses—and aid in the design of nanoparticles for a variety of purposes.
“In 2019 alone, the number of people who died due to antimicrobial resistance was 4.95 million. Even before COVID, which exacerbated the problem, studies showed that by 2050, the number of deaths from antibiotic resistance would be 10 million,” said Angela Violi, Arthur F. Thurnau Professor of mechanical engineering, and corresponding author of the study that made the conclusion of Natural Computational Science (“Domain-agnostic prediction of nanoscale interactions in proteins and nanoparticles”).
“In my ideal scenario, 20 or 30 years from now, I’d like—given any superbug—to be able to quickly produce the best nanoparticles that can treat it.”
Most of the work in the cell is done by proteins. Interaction sites on their surface can bring molecules together, break them down, and make other modifications—opening doors into cells, breaking down sugars to release energy, building structures to support cell groups, and more. If we can design drugs that target important proteins in bacteria and viruses without harming our own cells, it will enable humans to resist new and rapidly changing diseases.
The new model, called NeCLAS, uses machine learning—an AI technique that powers the virtual assistants on your smartphone and ChatGPT. But instead of learning to process language, it absorbed the known structural models of proteins and their interaction sites. From this information, he learned to extrapolate how proteins and nanoparticles might interact, predict the binding sites and possible bindings between them—and predict interactions between two proteins or two nanoparticles.
“Other models exist, but ours is the best for predicting interactions between proteins and nanoparticles,” said Paolo Elvati, UM associate research scientist in mechanical engineering.
AlphaFold, for example, is a widely used tool for predicting the 3D structure of proteins based on their building blocks, called amino acids. While this capacity is critical, it is only the beginning: Discovering how these proteins assemble into larger structures and designing practical nanoscale systems are the next steps.
“That’s where NeCLAS comes in,” said Jacob Saldinger, UM doctoral student in chemical engineering and first author of the study. “This goes beyond AlphaFold by showing how nanostructures will interact with one another, and not be limited to proteins. This allows researchers to understand the potential applications of nanoparticles and optimize their designs.”
The team tested three case studies that had additional data:
While many protein-protein models specify the amino acid as the smallest unit to be considered by the model, this does not work for nanoparticles. Instead, the team set the smallest feature size to be roughly the size of an amino acid, but then let the computer model decide where the boundaries between these minimum features are. The result is a representation of proteins and nanoparticles that looks like a collection of interconnected beads, providing more flexibility in exploring small-scale interactions.
“As well as being more general, NeCLAS also uses far less training data than AlphaFold. We only had 21 nanoparticles to look at, so we had to use the protein data in clever ways,” said Matt Raymond, UM doctoral student in electrical and computer engineering and study co-author.
Next, the team intends to explore biofilms and other microorganisms, including viruses.