When the virus hit the United States in force, Los Alamos did what many other organizations across the nation did: sent most of its workforce home. Apart from key national security personnel who continued to work onsite, Laboratory employees worked on laptops from home with various forms of connectivity software to keep their programs going. Of course, this approach had its limitations; for example, most scientists no longer had access to their laboratories or experiments. But far from giving up and watching soap operas in their pajamas, they instead answered a new call to serve the nation. Within just a few weeks, the Lab’s considerable supercomputer resources and expert personnel were allocated to national COVID-19 research efforts, and a rush of new internal research programs were variously proposed, approved, and underway.
SARS-CoV-2, the virus that causes the COVID-19 illness, is similar to other coronaviruses, such as those that cause SARS and MERS. Like them, it sprang upon the world suddenly, and relatively little could be known at the outset. There are questions about its history—how it emerged, over time and geography—that need to be answered. There are questions about its nature, such as how quickly it spreads and mutates, how often it kills, and whether or not it can re-infect. There are questions about its genomics—what genes or proteins might be targeted for vaccines and other treatments. There are questions of molecular structure—what proteins the virus constructs itself from, what shapes they take, and how their functions might be inhibited by different molecules—molecules that could serve as effective drugs, as long as they don’t introduce any toxicity themselves.
These sorts of questions require a tremendous amount of computational processing: examining one gene after another, checking every protein, screening vast numbers of existing and hypothetical drug molecules for both efficacy and safety. Whether by traditional large-scale data processing or advanced machine-learning techniques, high-performance computing enables rapid progress, perhaps condensing the timescale for the development of a vaccine or drug treatment. And supercomputer-based studies involving bioinformatics and molecular modeling have long been a specialty of Los Alamos National Laboratory. When the White House in mid-March announced a broad public-private coalition to support supercomputer research in the fight against COVID-19, Los Alamos—joined by a number of other national laboratories, NASA, and the likes of Amazon, Google, and IBM—proudly began sharing its considerable hardware and expertise with the entire national COVID-19 research enterprise.
Within just a few weeks, a rush of new research programs were variously proposed, approved, and underway.
In addition, while the country was largely closing down, Los Alamos scientists were massively spooling up a whole host of important research initiatives. Some are focused on the virus itself: its origin, its natural history, and its rate of evolution. Some are focused on direct vaccine development—against SARS-CoV-2 specifically or coronaviruses broadly, potentially to protect against future emergent pathogens. Others seek to help with testing and treatment activity, including investigating ways to increase the supply of necessary medical equipment (e.g., ventilators and face shields), such as 3D printing new equipment or sterilizing existing equipment for reuse. Still others are focused on epidemiology: forecasting the virus’s geographic and demographic spread and developing ready-to-use tools for informed, nearly real-time decision making in response to the evolving pattern of infection. Indeed, the Laboratory is partnering with Sandia, Argonne, and Oak Ridge national laboratories to produce a comprehensive, high-resolution pandemic model, integrating data collection and analysis to support policy makers. Los Alamos is also working with partner infectious-disease laboratories in countries such as Jordan, Uganda, the Republic of Georgia, and others for monthly exchanges of information on COVID-19’s genomics and molecular biology.
Several lines of research seek to analyze non-pharmaceutical mitigation strategies. How effective can we expect initial countermeasures, such as school closures and social distancing, to be over time? What about alternative strategies, such as more rigorous quarantines, with or without state-by-state variability? How will various scenarios affect rates of infection and death? Or the availability of key hospital resources, such as ventilators and healthy medical staff? And how can those resources be managed to optimal effect? Until drugs or vaccines are developed and distributed, answering these questions holds the greatest hope for minimizing the damage wreaked by SARS-CoV-2.
A global pandemic calls for two types of response. One is mass isolation, to protect individuals and slow transmission. The other is mass mobilization, to develop mitigation strategies, manage resources, and find a cure. When COVID-19 arrived, Los Alamos leapt into action on both fronts.
In-depth coverage of Los Alamos research to combat the COVID-19 crisis will appear in the next issue of 1663.