FLC | Argonne scientists detail role of APS in COVID-19 research

Argonne scientists detail role of APS in COVID-19 research

June 8, 2020

Imagine standing in Chicago, and firing at a foot-square target in Seattle. That's the precision with which scientists at Argonne National Laboratory are attacking the coronavirus on a daily basis.

"So people don't think of the Department of Energy as working in human health and biology," said the lab's Dr. Stephen Streiffer. "But we really have at the Department of Energy system, not just at Argonne but also at our sister laboratories around the country, great capabilities to study the structure of the virus, study the proteins that make up the structure of the virus, and use high performance computing to be able to simulate how different compounds might interact with those proteins."

Streiffer is director of Argonne's Advanced Photon Source, a one-kilometer circular structure where electrons are whirled to near the speed of light. In the process, they produce the world's brightest x-rays, which can be used to illuminate targets at submicroscopic levels, in this case, the two dozen or so proteins which make up COVID-19.

"The APS is this huge machine," Streiffer said. "And we're trying to hit protein crystals to determine their structure that may be on the order of about 20 microns or so, about a fifth the size of a human hair."

That's where that Chicago to Seattle shot comes in. Using the data from that analysis, Argonne scientists can map the proteins in detail. Then, they can run known compounds against that map, to try and determine which key might fit the lock.

"Maybe an actual better way to think about it," Streiffer said, "is instead of trying to find the key to that lock, really what we're trying to do is find the perfect super glue that can go into that lock and really gum it up so the virus quits working."

Because Argonne's THETA supercomputer works incredibly fast, so far they've screened about 4 billion combinations.

"We only need one," Streiffer says. "If we can find the perfect drug to interact with one protein, then we'd be able to solve the problem."

Elsewhere at Argonne, other scientists work with a computer model of the City of Chicago, down to every man, woman and child. That model, running on THETA, simulates their behaviors, their times at home or at work, and who they associate with.

"Everybody in Chicago is represented as what we call a software agent," said Argonne Distinguished Fellow Dr. Charles "Chick" Macal. "And we play that out over the course of an entire year, 8760 hours."

And THETA can run that Chicago year in 12 minutes.

"The model simulates the infection process, even simulates interactions of people at home, people going to essential businesses---grocery stores," Macal noted. "And what we're doing now in conjunction with the City of Chicago and the Chicago Department of Public Health, is to simulate re-opening scenarios."

The simulation is so detailed, Macal can ask his simulated population to do different things---such as wear masks and practice social distancing.

"And it's a very clear result," he said. "It's a very clear signal we're getting out of the data."

That signal? That the protective behaviors have been working.

"Those kinds of activities are extremely important and necessary for reducing transmission," he said. "Even if activities resume to their normal levels."

And, Macal said, that meant Illinois was on track for its re-opening schedule. However, the model has always been clear that if people backed off from protective behaviors, a second peak was likely in every scenario that was run.

"Re-opening is OK, in the sense of not increasing the number of infections over what they are on a daily basis, and not increasing the number of people who will become infected, and stay within the hospital capacities," he said. "We see that if people do practice protective behaviors, social distancing, and masks in particular, then yes we can get by without, for example, a second peak."