Sign up for our newsletter and get the latest HPC news and analysis.
Send me information from insideHPC:


Combating the Spread of Disease with HPC

In this special guest feature, Lance Farrell writes that, in this era of global air travel, NSF-funded researchers are using supercomputers to equip policy makers with the evidence needed to make fast, accurate choices in the event of an epidemic.

Strong as the weakest link. Policy makers will benefit from HPC simulations of disease spread. A flow chart of how differently-scaled models inform public policy is pictured. Courtesy Ashok Srinivasan.

Strong as the weakest link. Policy makers will benefit from HPC simulations of disease spread. A flow chart of how differently-scaled models inform public policy is pictured. Courtesy Ashok Srinivasan.

Welcome to the age of globalization. Transportation technology can take you anywhere in the world within hours. See the pyramids in the morning and be back stateside in time for dinner.

Of course that also means you can take some nasty viruses back home with you too. In a time of rapid transit, it’s not a matter of if transmission of infectious disease will occur  –  but when.

“I don’t wish to create a panic. However, with some bad luck, a single person could spread the Ebola infection to more than 15 persons,” admits Ashok Srinivasan. “The good news is that good airplane boarding procedures can reduce this substantially.”

Srinivasan, associate professor of computer science at Florida State University, leads a NSF-funded research team. His project, Viral Infection Propagation Through Air-Travel (VIPRA), with collaborators from Arizona State University and Embry-Riddle Aeronautical University, seeks to equip policy makers with scientific evidence for decision making.

Air travel is cited as the leading factor in spread of infectious diseases. Proliferation depends on many factors, including a person’s pathogenicity and infectivity, co-travelers’ susceptibility, duration of the flight, their seat location, and boarding and deplaning procedures.

But can a different type of a plane reduce transmission risk? Will travel limitations really stem the spread of infection?

Srinivasan wants to help decision makers answer these questions and react rationally.

To make policies that accurately reflect the spread of infectious diseases via air travel, officials need to understand spread at both the local flight scale and macro-scale across large geographic regions.

Presently, VIPRA uses fine-scale computational simulations to assess passenger movement while on an airplane. The initial focus is on Ebola, with other infectious diseases to follow. Future plans are to link with large-scale phylogeographic models to chart the path of an epidemic.

Courtesy Mikael Häggström; Centers for Disease Control and Prevention.

Courtesy Mikael Häggström; Centers for Disease Control and Prevention.

Using the ChainBuilder framework, VIPRA is developing a collaborative software infrastructure so that officials can make good use of these simulations. ChainBuilder is a web-based application that seamlessly integrates models and data from different domains, allowing easy manipulation of data to run what-if scenarios.

Linking fine-scaled details of transportation with the macro-scale can yield insights into possible global consequences of changes in procedures or policies at local scales,” says Srinivasan. “Such linking will be a major advance in virus surveillance, and is not feasible with current approaches.”

To perform these simulations, high-performance computers were crucial. “The quick turnaround time using such massive parallelism enabled us to refine out models iteratively in a much shorter time span than would have been feasible otherwise,” Srinivasan says.

Srinivasan’s team, which includes Anuj Mubayi, Sirish Namilae, Robert Pahle, and Matthew Scotch, looked to resources managed by the eXtreme Science and Engineering Discovery Environment (XSEDE). Blue Waters at the National Center for Supercomputing Applications, and Stampede at the Texas Advanced Computing Center figure prominently in Srinivasan’s work.

Srinivasan cites the NSF RAPID grant for quickly bringing together his team, and for developing computational infrastructure and techniques applicable to emerging epidemics.

Project VIPRA offers hope for science-based policies, where epidemic threats are handled rationally. Our research will enable decision makers to take fine-tuned decisions that mitigate the risk of an epidemic without taking drastic steps. It will also help reassure the public that action is being taken to deal with a potential epidemic.”

With a rising world population and our expanding web of technology, it seems our planet is becoming a smaller place. High-speed networks bring the world into our living room, and modern air travel means the rest of the world is just a hop, skip, and a jump away.

Infectious diseases are deadly serious business, but NSF-funded researchers are using supercomputers to keep us safe from their spread.

XSEDE16 will be held in Miami from July 17-21.

Source: Science Node under the Creative Commons license.

Sign up for our insideHPC Newsletter

Resource Links: