As COVID-19 sweeps the globe to devastating effect, supercomputers around the world are spinning up to fight back by working on diagnosis, epidemiology, treatment and vaccine development. Now, Lawrence Livermore National Laboratory (LLNL) has announced that it has leveraged high-performance computing to help investigate candidates for new antibodies and drugs to fight COVID-19. In an article by Jeremy Thomas, LLNL highlighted this and other COVID-19 computing research under its umbrella.

LLNL’s COVID-19 response has been deeply interdisciplinary. First to the plate was Adam Zemla, a structural biologist and mathematician at LLNL, who led a team of researchers who produced a predicted 3D protein structure of COVID-19. That predicted structure – which has now been verified by identification of the crystal structure of the protein – was used by more than a dozen outside research groups. 

Meanwhile, within LLNL, a team led by Daniel Faissol (a group leader in operations research and system analysis) and Thomas Desautels (a staff scientist) took that structure and ran with it. Using the prediction and knowledge of antibodies that have proved effective against SARS, the research team used a pair of HPC clusters to apply AI to antibody analysis simulations. According to LLNL, the simulations are the first to integrate machine learning, experimental data, structural biology, bioinformatic modeling and molecular simulations to design possible antibody candidates. 

“Our approach, while still being developed, is aimed at designing high-quality antibody therapeutics or vaccines in extremely rapid timescales for scenarios in which waiting for many rounds of time-consuming experimental steps is not an option,” Faissol said. “Experimental data and structural bioinformatics are important components to enable high-quality predictions, but integrating machine learning and molecular simulations on HPC are key to enabling the speed and scalability we need to search and evaluate huge numbers of possible antibody designs.”

The result? 10³⁹ possible antibodies narrowed down to 20.

LLNL hasn’t stopped there. Another group, led by Felice Lightstone (a senior scientist) and (a researcher) examined 26 million molecules, testing how they interacted with COVID-19’s key proteins. To examine the molecules, the team applied the entire Quartz cluster (an Intel Xeon-based cluster delivering 3.25 peak petaflops).

“This is the first step toward finding a new antiviral,” Lightstone said. “We developed a whole pipeline for drug design and plan to continue in the coming weeks, ending with experimental testing of the predicted molecules. This should speed up the drug design process.”

LLNL sees the application of its supercomputing power as a direct and crucial validation of its resources and research programs – and the amount of LLNL computational resources dedicated to COVID-19 seems likely to grow.

“For several decades, the Laboratory has been at the forefront of protecting the country against biological threats of any type,” said Dave Rakestraw, the senior science adviser who is coordinating LLNL’s technical response to COVID-19. “We’ve been putting a large amount of focus for the last six years on using the computational resources at LLNL to try to accelerate the timescales for developing a response to an emerging biological threat. We’ve done that […] by using our extensive computational capabilities (staff and computer infrastructure) and developing partnerships with universities, drug companies and tech companies. That effort has put us in a position where we have tools now that are applicable to helping with the current response.”

“The Laboratory anticipated this kind of situation in pursuing a predictive biology initiative,” added Shankar Sundaram, director of LLNL’s Center for Bioengineering. “The reason we were able to jump onto this quickly was not just because we had the capabilities, but because we’ve been thinking about these scenarios for a long time.”