Modeling of large macromolecular systems at extra-long time scales using AI

Modeling of large macromolecular systems at extra-long time scales using AI is sponsored by National Science Foundation (NSF). This project focuses on modeling molecular systems using artificial intelligence to expand scientific knowledge of protein structures and their assembly in crowded, cell-like environments.

Featured news and headlines | KU News KU Communicator Resources When Experts Attack! podcast $1. 5 million NSF grant will support KU research modeling molecular systems using AI LAWRENCE — University of Kansas researcher Ilya Vakser is leading a new National Science Foundation-funded project on modeling molecular systems using artificial intelligence.

Ilya Vakser Vakser, professor of molecular biosciences and director of the KU Computational Biology Program, will collaborate on the three-year, $1. 5 million grant with researchers at Stony Brook University and Stowers Institute.

By splitting the funding among these three teams, Vakser said, the project combines highly complementary areas of expertise on structure-based modeling of molecular interactions, theoretical foundation of reinforcement learning — a major direction in AI — and experimental approaches to protein aggregation.

The grant, titled “Modeling of large macromolecular systems at extra-long time scales using AI,” will allow Vakser and colleagues to expand scientific knowledge of protein structures and their assembly in the crowded, cell-like environments.

“This research is essential for scientists to better understand the work of the molecular machines that enable biological processes, allowing for the increased ability to modulate these processes when appropriate,” Vakser said. The research will generate the energy landscape of the macromolecular system as the foundation of the simulation approach, which samples the landscape in space and time, according to the research team.

The accuracy of the energy landscape will be significantly improved by application of the AI-based predictions. The modeling protocol will be radically advanced to a new AI-driven paradigm, according to the researchers.

Instead of explicitly simulating the entire trajectory of the system, the project will put forward AI methods that learn from the early segments of simulations and from the system parameters to model the long-term behavior. The emergent protein behavior will be investigated for the eye lens proteome. Crystallin protein oligomerization competes with aggregation, and how it is autonomously regulated is poorly understood.

The project puts forward a hypothesis that the lens crystallins likely operate through an entropic buffering mechanism. The aggregates promote their own growth by excluding volume that would otherwise be accessible to monomers, which creates an entropic drive for further aggregation.

The modeling approach, combined with wet-lab experiments, will provide a unique opportunity to test this hypothesis and to explore fundamental biology of protein homeostasis, according to the researchers. The grant also will involve graduate and undergraduate students in projects that apply AI to key problems in molecular biology.

“This education experience will allow for better career prospects in the current time of technological revolution,” Vakser said. He said the research will lead to the radical expansion of the existing approaches to modeling macromolecular processes in cells by incorporating the AI techniques.

“Thus the project will open up an uncharted territory for research, with unprecedented opportunities for life sciences that will have significant impacts on medicine, biotechnology, agriculture and beyond,” Vakser said.