The number of people donating their PC’s computing power to help develop new drug therapies against the coronavirus has skyrocketed.
Crowdsourcing project [email protected] specializes in trying to find weaknesses in diseases by simulating them using the public’s spare computing resources. The project originally had around 30,000 active users before the coronavirus pandemic started. But in the past two weeks, the figure has shot up to more than 400,000 users.
The 1,200 percent increase in volunteers was mentioned in a Reddit Ask Me Anything the project’s leaders held on Thursday. In late February, [email protected] announced it would start studying the coronavirus strain, COVID-19 in the hopes of creating an accurate model of the disease and its vulnerabilities.
Since then, companies including Nvidia have called on consumers to donate their PC’s unused computing resources to the cause. To do so, download and install the [email protected] app. To contribute to the coronavirus research, click on “any disease” when selecting which illness you want to help fight.
The 400,000-plus volunteers means the [email protected] project has access to 474 petaflops in computing power. That’s over twice the power of the world’s fastest supercomputer, Summit, which has also been deployed to aid researchers in studying COVID-19.
But how does simulating a virus actually help? [email protected] hopes to understand how the disease’s viral proteins behave and bind to a victim’s lung cell. These proteins are constantly folding into different microscopic structures. As a result, taking static pictures of the coronavirus in action isn’t enough to fully dissect the illness. “Proteins actually move, and a lot of the information about that that is not there in static pictures needs to be collected from simulations,” Rafal Wiewiora, a researcher with the project, said during the Reddit AMA.
Specifically, the project is focusing on finding holes or pockets in COVID-19’s viral proteins that doctors can “squeeze a drug molecule” inside to neutralize the disease, Wiewiora said. The simulations will also calculate how strongly a particular drug molecule binds to a viral protein All the results will be shared with the medical community and pharmaceutical companies to help them refine a drug for market.
Thanks to the surge in volunteers, [email protected] has been able to run an unprecedented number of protein simulations, and at a faster rate. “Personally, the turnaround speed from setting up a project to getting useful hypotheses from the data and making decisions on what to do next, has improved immensely — I can now run a protein, come back in a week and already know what the next step is, this would take a month to a few months before,” Wiewiora said.
“We have been beating records in simulation speed and data amount generated for a long time, but what is happening now is an order of magnitude more — really a milestone in distributed computing,” he added.