AsianScientist (Mar. 18, 2022) – Almost two decades ago, countries across Asia were imperiled by a contagious and deadly disease caused by a coronavirus. In a matter of weeks, the severe acute respiratory syndrome (SARS) outbreak crossed international boundaries, resulting in thousands being infected and a death toll of nearly 800, coupled with severe pressure on Asia’s interconnected systems.
In early 2020, the same systems, including healthcare, research, transportation and logistics, were once again tested by a coronavirus. Within a year, SARS-CoV-2, the virus behind COVID-19, had spread across the continent and the world, wreaking havoc on almost every economy.
While nations may have responded differently to the current pandemic, one common truth has stood out—those that take the opportunity to adapt and transform will emerge stronger.
Responding quickly in the early days of the COVID-19 outbreak, scientists around the globe began tapping into the massive power of high performance computing (HPC). They were able to process large volumes of COVID-19 data and simulate biological and chemical processes by running complex mathematical models on supercomputers.
These simulations not only advanced our understanding of the evolving virus and how it causes human infection, but also paved the way for the rapid development of interventions against it. Read on to find out about how organizations, governments and universities in Asia have harnessed supercomputer solutions to power throughout the pandemic.
Supercharging test kit development
At a virtual press conference in March 2020, the director-general of the World Health Organization, Dr. Tedros Adhanom Ghebreyesus, gave clear advice to all nations, “Test, test, test. Test every suspected case.”
But to roll out testing en masse, a country first needs test kits. Heeding the call, Seoul-based biotech company Seegene used their supercomputer to develop a test kit within a matter of weeks, rolling them out even before other countries confirmed their first cases of the new virus. The test kit, based on the real-time reverse transcription polymerase chain reaction (RT-PCR), could detect the presence of viral genes in suspected patient samples.
Despite the lack of physical samples from patients on hand, the team used artificial intelligence to analyze the molecular blueprint of SARS-CoV-2 and assemble the required reagents that would react with the virus’ genetic material to create a highly sensitive test. Without the supercomputer, it would have taken the team two to three months to develop such a test, Dr. Chun Jong-yoon, founder and CEO of Seegene, told CNN.
Mission: Infection Detection
Diagnosing COVID-19 currently relies on RT-PCR testing, considered the gold standard for detecting viral genetic material from clinical samples. If initial RT-PCR testing delivers inconclusive results, though, chest computed tomography (CT), similar to X-ray scans, may be used instead.
However, doctors have to spend a considerable amount of time—at least 15 minutes—assessing CT images, proving a major bottleneck when these healthcare experts are in such short supply.
To bridge this gap, researchers at Tsinghua University in China devised an HPC-enabled artificial intelligence (AI) tool to perform the analysis in a significantly shorter timeframe, taking only a few minutes or even mere seconds. The supercomputer that was called in to help was the Elastic High-Performance Computing (E-HPC) Solution for Life Sciences by Chinese e-commerce and cloud giant Alibaba.
The E-HPC includes four Intel central processing unit-based configurations and one similar configuration with additional Tesla P100 graphical processing units, all with network speeds ranging from 10 to 25 Gbps. With such processing power, the E-HPC easily accommodated the algorithms and data requirements, allowing the AI program to quickly learn to pinpoint COVID-19 diagnoses from CT scans.
Beyond diagnosis, researchers at Sun Yat-Sen University also leveraged E-HPC’s ability to accelerate data transfers to gain more detailed insights into the virus’ evolution. Meanwhile, a Peking University team harnessed E-HPC’s resources to speed up their molecular docking tests, which analyze how molecules fit together. Given their vast computational power, such supercomputers are valuable in advancing the discovery of antibody and drug candidates to target the virus where it’s most vulnerable.
Unlike their conventional counterparts, HPC-powered models can look at tons of different molecules at once, simulating how they might latch onto and disable SARS-CoV-2’s spike protein, the key to entering and infecting human cells.