While not the novel coronavirus that is now sweeping across the world, the 2009 H1N1 flu pandemic (pH1N1) infected up to 21 percent of the global population and killed over 200,000 people. Now, a team of researchers from the University of California San Diego have applied supercomputing power to pH1N1 simulations in order to better understand that pandemic, manage its annual recurrences – and prepare for the next one.

The UC San Diego team built an all-atom, solvated and experimentally based integrative model of pH1N1, using this simulation to examine two binding sites on the flu’s viral proteins. While most viral simulations simplify in order to precisely model one target area for a drug, this simulation broke new ground by modeling the entire viral envelope of pH1N1 – over 160 million atoms – without sacrificing detail.

“Even just building this system from the ground up was quite challenging as it required us to integrate different types of experimental data, at different resolutions,” said Rommie Amaro, the UC San Diego professor of chemistry and biochemistry who led the study, in an interview with UC San Diego. “This model itself was useful to understand the physical arrangement of molecules in the virus, which had never been done before at this level of detail.”

To successfully run the mesoscale simulation of pH1N1 – one of the largest biological systems ever to be simulated on this level – the researchers turned to the Blue Waters supercomputer hosted by the National Center for Supercomputing Applications (NCSA) at the University of Illinois at Urbana-Champaign. Blue Waters, now in its final year of operations, is a hybrid CPU-GPU Cray system with more than 1.5 petabytes of memory and approximately 26 petabytes of storage that delivers around 13 peak petaflops.

A video of the simulation, courtesy of Lorenzo Casalino, UC San Diego.

The powerful simulation allowed researchers to better understand how flu proteins interacted with one another on the surface of the virus. They also used the many protein copies throughout the simulation to conduct a statistical analysis of viral protein movement. “This was useful because it allowed us, for the first time, to characterize how quickly one particular loop known as the ‘150-loop’ opened and closed,” Amaro explained. “We care about the motion of this loop because the loop is right up next to where drugs, like Tamiflu, bind to deactivate the protein.”

Crucially, the study may have opened the door for new anti-influenza therapeutics. “Our study also provides new evidence that an often-overlooked so-called ‘secondary site’ may be the first place the natural substrate of the flu binds,” Amaro said. “Thus a novel viable therapeutic strategy may be to design molecules that effectively block that site.”

“We are thrilled Blue Waters, both the computer and the support staff, was able to contribute to this groundbreaking success,” said William Kramer, director of Blue Waters. “To our knowledge, a molecular dynamics simulation on so grand a scale has never been attempted, let alone completed.”

Funding for the work was provided in part by the National Institutes of Health and the National Science Foundation.

About the research 

The research discussed in this article was published as “Mesoscale All-Atom Influenza Virus Simulations Suggest New Substrate Binding Mechanism” in the February 2020 issue of ACS Central Science. It was written by Jacob D. Durrant, Sarah E. Kochanek, Lorenzo Casalino, Pek U leong, Abigail C. Dommer and Rommie E. Amaro and can be accessed here.

To read the article by UC San Diego’s Cythia Dillon discussing this research, click here.