Coronavirus graphic

World’s fastest supercomputer identifies coronavirus drug candidates

Image credit: University of Tennessee/Oak Ridge National Laboratory

IBM AC922 Summit, the first supercomputer capable of reaching a quintillion operations per second, has identified 77 compounds with the potential to be used to fight the coronavirus responsible for the ongoing pandemic.

Summit, which is based at Tennessee’s Oak Ridge National Laboratory, is used by researchers to perform extremely complex tasks – such as climate simulations and identifying risk factors for certain diseases – within a reasonable timeframe.

Now, scientists have used the machine to identify several dozen drug compounds that could have the potential to fight the ongoing Covid-19 pandemic, which is caused by a novel coronavirus, by stopping the virus from infecting host cells.

Coronaviruses infect host cells through what is essentially an injection of genetic material via a 'spike' protein (S-protein). Micholas Smith, a postdoctoral researcher at the University of Tennessee, used a 3D model of the spike protein of the new coronavirus interacting with the ACE2 receptor to identify molecules that bind to the virus’s S-protein, essentially 'filling' the interface through which the virus infects a host cell.

“Given that the outbreak of SARS-Cov-2 may quickly become a global challenge, it would be of great benefit to identify and repurpose already well-characterised small molecules and approved drugs for use in combating the disease,” Smith and his colleagues wrote in their report.

Using Summit to perform simulations of how the S-protein would respond to more than 8,000 different molecules, the researchers were able to narrow down a list of thousands of compounds to a ranked list of 77 which have the potential to stop the virus infecting host cells. Seven of these, which are already available with regulatory approval or have had multiple previous studies with promising results – were highlighted as the strongest possibilities.

The researchers plan to run the simulations on the supercomputer a second time, using a more accurate model of the coronavirus’s S-protein published even more recently in Science (having based their model on research published in January).

“Summit was needed to rapidly get the simulation results we needed. It took us a day or two whereas it would have taken months on a normal computer,” said Jeremy Smith, director of the Oak Ridge National Laboratory Center for Molecular Biophysics.

“Our results don’t mean that we have found a cure or treatment for the coronavirus. We are very hopeful, though, that our computational findings will both inform future studies and provide a framework that experimentalists will use to further investigate these compounds. Only then will we know whether any of them exhibit the characteristics needed to mitigate this virus.”

Experimental studies will be the next crucial step in demonstrating which compounds are most effective at defending against the virus. With the coronavirus pandemic continuing to spread rapidly around the world, forcing governments to take measures unprecedented in peacetime to contain it, there is intense pressure on researchers to develop a safe, effective treatment for the virus in a matter of months rather than the decade or so it would typically take for such a treatment to be developed, trialled, and approved.

Sign up to the E&T News e-mail to get great stories like this delivered to your inbox every day.

Recent articles