DOE Launches ‘Virtual’ National Lab to Counter Coronavirus

This article is adapted from an April 28 post by Physics Today, which is also published by the American Institute of Physics.

To coordinate its contributions to the global effort to combat the SARS-CoV-2 virus, on April 16 the Department of Energy announced its creation of a “National Virtual Biotechnology Laboratory” (NVBL) that will serve as a clearinghouse for access to experts and equipment across the entire lab system.

Facilities available to researchers include light and neutron sources, nanoscale science centers, sequencing and bio-characterization facilities, and high-performance computing assets. The virtual lab’s computing, diagnostics, epidemiology, and other teams are also tied into a national task force run by the Federal Emergency Management Agency and the Department of Health and Human Services (HHS).

Under the CARES Act, signed into law on March 27, national laboratories administered by the DOE Office of Science and National Nuclear Security Administration received $99.5 million in immediate funding for coronavirus research and user facility operations. Physics Today spoke with Chris Fall, director of the DOE Office of Science, about the goals of the virtual lab construct.

Physics Today: How will the NVBL work? If I am a researcher who needs research tools or wants to develop models of the outbreak, what should I do?

Fall: If you are doing the kind of work that we do or are doing drug discovery at a major company, you are probably already working with us. There are only so many synchrotrons and free-electron lasers. The concept behind the NVBL is that many of our national labs have biotechnology capabilities well beyond diagnostics, computing, light sources, and neutron sources. They have substantial molecular biology expertise and capabilities.

The labs like to compete for funds. Because of the crisis, they’ve come together in a fundamentally unique way through the NVBL. Under the virtual lab model, we have one front door for everything COVID related. This includes ideas coming in from the outside and ideas coming in from the labs themselves. In the new commons, if you will, all the labs can say, “Here’s how I can help, here’s the technology, here’s the special equipment or the facilities we have.” Proposals on a topic are evaluated and prioritized by a steering group composed of representatives from all the labs and senior representatives from DOE. The research selected will be packaged and funded as a unit.

The NVBL also is prototyping what it might be like if we want to draw from the technology capability of all the labs under a one-front-door model. In the absence of a national lab for biotechnology and biosecurity, this virtual biotechnology laboratory concept may serve as a model for a construct for general biotechnology/biosecurity/bioeconomy work that could persist longer-term after the COVID-19 crisis subsides.

PT: Some of the light sources in Europe are now screening compounds to see what might inhibit SARS-CoV-2 proteins. In the U.S., we seem to be focusing on computational screening. Is one approach better than the other?

Fall: They are complementary. There are a ton of compounds and a ton of places where compounds could bind to a protein. The computational work, whether traditional or AI-enabled, helps us winnow down the landscape of possibilities to things we will experiment on. We don’t go from computers to injecting people directly. There’s always going to be an experimental verification step. Our light sources are still operating. We’ve kept open anything we think can be used in the fight against COVID.

PT: The NVBL announcement says that DOE labs’ facilities and trained personnel “can be deployed to help address the rising surge in clinical samples.” Are the labs going to be doing coronavirus testing?

Fall: We are doing some testing at our labs. We’re building a testing capacity mainly to test our own workforce. That’s being done by our occupational health people, not the NVBL. We’re not building a LabCorp. That’s not our mission, and it’s not something Congress would allow. With the NVBL, we are trying to improve testing and make it more efficient. There are a number of issues with testing, such as the availability of reagents and the fact that PCR [polymerase chain reaction] is generally the only kind of test available. We are looking at alternative testing methodologies and other reagents.

PT: Are there any successes to report so far from virtual collaborations?

Fall: We’ve just gotten started. But already our project on pandemic modeling and analysis has shown that the mobility change in the U.S. in response to COVID-19 has been dramatic and preceded most state-level stay-at-home orders. It found that 95% of U.S. residents are living in counties where typical mobility has been cut in half, and nearly half of U.S. residents are living in counties where typical mobility has dropped by 95%. These data provide valuable insight for planning and formulating a COVID-19 response.

PT: Another stated mission for the NVBL is addressing supply-chain bottlenecks. Are the labs going to be manufacturing masks and ventilators?

Fall: We’re not looking to be a supplier, but we are looking at how to help fix the supply chain for masks and ventilators. We do a lot of manufacturing technology work, including very sophisticated 3D printing. The problem isn’t printing a 3D part; you can get a printer for your desk. The secret is producing parts that you know will work right and will last. That requires a deep understanding of materials science and the melting process the 3D printers use. Materials science is one of our core missions.

One of our technology partners, Battelle, has been in the news for its technology to chemically sterilize and recycle things like N95 masks. There are other ways to do that using light and radiation, things we are pretty good at. So we are looking at other ways of recycling personal protective equipment to help reduce the bottleneck. Under the NVBL construct, that role is tied into the supply chain team within the vice president’s task force. They ask us questions, and we see if we can provide answers.

PT: Epidemiology is another stated NVBL function. Do the labs have resident expertise in this, or would epidemiologists at other agencies and in academia come to the labs for help?

Fall: The labs absolutely have epidemiology expertise, stemming partially from DOE’s national security mission. Some of the labs are expert at tracking people and things around the world that are relevant to national security. We bring some pretty unique things to epidemiology, not least the scale and capabilities of our computers.

There also is an ongoing epidemiology research program funded by HHS in a number of the labs. COVID is a problem that’s different from what they might normally be looking at, but the tools, the software, expertise, and data sources are things we already have available.

PT: Does epidemiological modeling require the enormous computing power that the labs have?

Fall: If you are doing simple, one-dimensional epidemiological models, numbers over time, it doesn’t take a big computer to do that. But if you want epidemiological modeling that includes overlaid details such as a geographical grid, detailed differences in local populations, or complicated and detailed travel patterns, it starts to be a computer problem. We do all sorts of overlaying very well. They have fantastic epidemiologists at HHS and the CDC [Centers for Disease Control and Prevention], but they don’t make spatial models that include all these other factors. That’s a specialty of DOE.

PT: Another capability touted in the NVBL announcement is “specialized facilities consisting of a multi-tier architecture facilitating a private cloud environment.” What does that mean?

Fall: We are helping lead the COVID-19 High Performance Computing Consortium for the White House. That consists of our computers, NSF’s, NASA’s, and a number of big iron [very expensive and powerful] computers from IBM, Microsoft, Amazon, Google, and other industry and academic institutions. A lot of people want to get their ideas onto a supercomputer. Those ideas are being submitted to a central clearinghouse, where they are quickly evaluated by a team of experts. The ones that are most promising are farmed out to the consortium members, who say, “You take this one, I’ll take that one, this one is right for your kind of computer.” That’s what we are talking about in the multi-tier computing and cloud services.

The other thing we do at DOE is operate protected data enclaves. We know how to take care of special data, not just national security data but also data that are sensitive because they’re proprietary or include personally identified information. We know how to bring users into our facilities, let them compute, and even let other people compute with those data without sharing secrets. That’s particularly applicable in this case, where there will be patient data. You want to be able to mine the data without revealing identities.