‘Breakthrough’ study offers powerful computational modeling approach for cellular simulations

A landmark report from the University of Kansas published this week in the Proceedings of the National Academy of Sciences proposes a new technique for modeling molecular life with computers.

According to lead author Ilya Vakser, director of the Computational Biology Program and Center for Computational Biology and Professor of Molecular Biosciences at KU, the investigation of computational modeling of life processes is a major step towards creating a simulation of a living cell at atomic resolution. . This advance promises new insights into the basic biology of a cell, as well as faster and more accurate treatment of human diseases.

“It’s about tens or hundreds of thousands of times faster than existing atomic resolution techniques,” Vakser said. “This provides unprecedented opportunities to characterize physiological mechanisms that are now far beyond the scope of computer modelling, to better understand cellular mechanisms, and to use this knowledge to improve our ability to treat disease.”

Until now, a major obstacle to computer modeling of cells has been approaching proteins and their interactions that are at the heart of cellular processes. To date, established techniques for modeling protein interactions have relied on either “protein docking” or “molecular simulation”.

According to the researchers, both approaches have advantages and disadvantages. Although protein docking algorithms are great for sampling spatial coordinates, they don’t take into account the “time coordinate” or dynamics of protein interactions. On the other hand, molecular simulations model the dynamics well, but these simulations are too slow or at low resolution.

“Our proof-of-concept study bridges the two modeling methodologies, developing an approach that can achieve unprecedented simulation time scales at all-atom resolution,” the authors wrote.

Vakser’s collaborators on the paper were Sergei Grudinin of the University of Grenoble Alpes in France; Eric Deeds of the University of California-Los Angeles; KU PhD student Nathan Jenkins and Petras Kundrotas, assistant research professor in KU’s Computational Biology program.

After conceptualizing how best to combine the advantages of both protein modeling approaches, the team developed and coded an algorithm to drive the new simulation.

“The most difficult challenge was to develop the algorithm that adequately reflects the simple basic idea of ​​the approach,” Vakser said.

But once they have made that breakthrough, they could start validating the new procedure.

“The paradigm was easy – a stroke of clarity,” Vakser said. “Existing simulation approaches spend most of the computational time traveling through low-probability – or high-energy – areas of the system. We all know where those areas are. Instead, the idea was to ‘sample or travel only in high-probability, low-energy areas, and skip low-probability ones by estimating transition rates between high-probability states. The paradigm is as old as biomolecular modeling itself and has been widely used since the dawn of the modeling era decades ago.”

But Vakser said that until his team’s new paper, the approach had not been applied to the kinetics of protein interactions in the cellular environment, the focus of their study.

“Because there are far fewer high-probability states than low-probability states, this gave us a huge boost in computational speed — tens to hundreds of thousands of times,” Vakser said. “This was done with no apparent loss of accuracy. It can be argued that accuracy has been gained, as the simulation protocol is based on ‘docking’ techniques, which are specifically designed to characterize protein assemblies.”

The KU researcher said his cellular simulation method could be deployed to research human health and treat disease with a new level of precision.

“The approach can be used to investigate molecular pathways underlying disease mechanisms,” Vakser said. “It can be used to determine the harmful effects of genetic mutations by the altered patterns of protein associations – genetic mutations cause changes in protein structure, which in turn affect protein association. Or it could be used to identify targets for drug design by detecting critical elements in protein association patterns.”

According to Vakser, the new simulation technique offers many promising avenues of research to explore in the future.

“Among them is adapting the approach to protein interactions with nucleic acids, RNA and DNA,” he said. “Furthermore, we would like to take into account the flexibility of molecular shapes, correlate with the rapidly developing spectrum of experimental studies of the cellular environment, and apply the procedure to a model of a real cell – with its real molecular components grouped together. “

Comments are closed.