Researchers use ALCF resources to model the spread of COVID-19

Occupancy patterns in the Chicago space as generated utilizing the Argonne CityCOVID model. Credit: Argonne National Laboratory

With COVID-19 drastically altering day by day life for individuals throughout the planet, the U.S. Department of Energy’s (DOE) Argonne National Laboratory has moved shortly to be part of the international combat towards the pandemic. Among the laboratory’s strongest resources for scientific analysis is the supercomputer Theta, housed at the Argonne Leadership Computing Facility (ALCF), a DOE Office of Science User Facility. Over 250 nodes on the machine had been instantly reserved for multipronged analysis into the illness.

Led by Argonne computational scientist Jonathan Ozik and Argonne Distinguished Fellow Charles (Chick) Macal, one of these branches of analysis oversees the improvement of epidemiological fashions to simulate the spread of COVID-19 all through the inhabitants.

The fashions are city-scale simulations of Chicago, populated with slightly below three million brokers that characterize people going about their day by day schedules and navigating some 1.2 million websites (homes, faculties, workplaces, and so forth) that every current potentialities for them to meet, or colocate—that’s, potentialities for publicity. Following publicity, an agent can grow to be contaminated in a extreme method, relying on an agent’s profile, which incorporates age traits. A sure quantity of the contaminated brokers then perish.

These fashions—operating for a simulated yr—are revised and improved every day, in accordance with the most up-to-date knowledge and knowledge. These updates are shifting towards a totally automated workflow.

“The workflow ingests up to date epidemiological knowledge—for example, that revealed day by day by the Chicago Department of Public Health—which function empirical goal trajectories. By evaluating these with outputs generated from ensemble runs, we’re ready to estimate the pandemic’s underlying parameters,” Ozik mentioned. “It is these calibrated parameters that enable us to run different scenarios with the model.”

Researchers use ALCF resources to model the spread of COVID-19
Agent mobility patterns, as generated utilizing the Argonne CityCOVID model. Credit: Argonne National Laboratory

“This is the most detailed granular simulation of COVID-19 that exists right now in terms of modeling individuals who could be in various disease states, including infectious or hospitalized,” Macal mentioned.

The fashions pursue strains of inquiry that shall be acquainted to anybody following the virus in the information media—for instance, the distinction in consequence yielded by implementing social distancing measures for nonetheless many extra days or even weeks.

“What are good ways to ease off the social distancing measures?” Ozik requested. “Everybody’s interested in that for very obvious reasons, but we don’t want to do something that will just create another calamity a few months down the road.”

The vital computational calls for of the undertaking end result from the fashions’ stochastic (randomly decided) parts, which govern the underlying uncertainties and parameters of the simulation. These parameters govern agent behaviors, in addition to illness development dynamics and transmissibility. Within the model, transmissibility encapsulates the probability {that a} prone agent is contaminated, primarily based on the quantity of time that two brokers spend collectively.

Researchers use ALCF resources to model the spread of COVID-19
Endogenous contact networks, as generated utilizing the Argonne CityCOVID model. Credit: Argonne National Laboratory

“With this model, you have potentially many people interacting in many different ways: some might be infected, some might be susceptible, and they mix in different proportions in a variety of different locations—there are different locations like schools and workplaces where very different parts of the population interface,” Ozik defined. “The multitude of possibilities the model presents make it quite qualitatively different from—and quantitatively more complex than—a statistical model or more simplified compartmental models, which are much faster to run.”

With optimization help from ALCF workers, simulation runs on Theta have utilized greater than 800 nodes directly. As half of the automated workflow, following these simulation runs, output knowledge are transferred to Petrel (a service supplied by Argonne and Globus, a University of Chicago-run non-profit dedicated to knowledge administration) for archival storage and post-processing; this post-processing is accomplished on Bebop, a high-performance computing cluster operated by Argonne’s Laboratory Computing Resource Center that the group additionally leverages for runs.

“It’s the big picture that we’re trying to capture with these simulations,” Macal mentioned. “How can we innovate and contribute by generating information unavailable anywhere else? We want to have an impact on the decisions that are being made about social distancing and opening society back up.”

Expanding the limits of personalized medicine with high-performance computing

Researchers use ALCF resources to model the spread of COVID-19 (2020, May 28)
retrieved 10 June 2020

This doc is topic to copyright. Apart from any truthful dealing for the function of personal examine or analysis, no
half could also be reproduced with out the written permission. The content material is supplied for data functions solely.

Source link

Leave a Reply

Your email address will not be published. Required fields are marked *