Updated: May 19, 2020
Author: Sohail Merchant, MD.
How a supercomputer and AI are taking the fight to COVID-19.
The Pandemic’s Rise
As the world’s fight towards reducing the number of COVID-19 cases rages on, it is clear that the disease is spreading faster than initially anticipated. Thus, our need for a vaccine or a combination of drugs that can assist patients in their full recovery is more paramount.
Researchers like Vincent et al. attempted to alert the world about this impending danger long before it became a crisis. In 2007, their warning bells fell on mostly deaf ears. Their research work, titled “Severe Acute Respiratory Syndrome Coronavirus as an Agent of Emerging and Reemerging Infection,” characterized the habit of the consumption of horse bats and other exotic mammals in Southern China as a ticking time bomb. They cautioned that the animals carried viruses like SARS-CoV, which might have precipitated the outbreak of SARS in 2003, and quite possibly, our current COVID-19 global emergency.
With over 2 million people testing positively for the novel coronavirus and still counting, many regions of the world are under complete lockdown as we attempt to stifle the spread. There is also a frantic effort aimed at creating a vaccine and/or identifying the right combination of drugs that might combat the pandemic’s advancement.
Additionally, our global movement to limit person-to-person interaction has spawned a large market for AI-based systems and robots that can replace humans with little to no supervision. See our article about them here.Innovative technology has been called upon to deliver crucial items, identify sick people, reduce healthcare worker burnout, and even pass out hand sanitizer. Scientists are also leveraging AI and supercomputers to uncover those drugs and compounds that can help us resolve our harrowing COVID-19 situation more quickly and safely.
AI, Supercomputers, and Flattening that Curve
A research work published in February 2020 explains how SUMMIT, the world’s fastest supercomputer, identified 77 different compounds that might limit the spread of COVID-19. In their research, Smith et al. (2020) identified that SARS coronavirus infections are known to be closely related to interactions between the viral spike protein (S-protein) and specific human host receptors like the Angiotensin-converting enzyme 2 (ACE2) receptor. Spike proteins are large transmembrane proteins that contain roughly 1160 amino acids for avian infectious bronchitis virus and about 1400 amino acids for feline coronavirus.
Before going on to give their list of compounds that can inhibit the interaction between coronaviruses and the receptor proteins, Smith et al. (2020) submitted that the interplay of the SARS-CoV-2 S-protein and the human ACE2 complex is indeed what facilitates the COVID-19 infection. Accordingly, that reciprocity is a logical target. They also advise that traditional drug development in the United States can take up to 15 years; time is not on our side if we must abide by the usual rules. As of mid-April 2020, the worldwide death toll stands at more than 157,000.
The story behind the creation and operation of SUMMIT is fascinating. Developed by IBM under March 2014 directives from the U.S. Department of Energy, it was slated to be 5 to 10 times faster than Titan, the world’s fastest most powerful computer at that time. The development of SUMMIT (and Sierra, the world’s other fastest supercomputer) took over four years and featured a system with brains comprised of artificial intelligence and deep learning, while 125 petaflops (Sierra) and 200 petaflops (SUMMIT) respectively act as their brawn. When the dust settled and they were complete, it was discovered that they were one million times better than the fastest laptop and powerful enough to sift through millions of variables as they created models and simulations.
SUMMIT was literally designed to tackle the world’s challenges: to combat cancer, to identify better materials like semiconductors, batteries, and building materials, and to accelerate how we understand diseases.
According to the research findings, COVID-19 first affects the outer cells of the human body before gaining entry into the respiratory system where, as we know, it wreaks its havoc. Out of the 77 possibly useful compounds and drug targets identified by SUMMIT, Smith et al. (2020) singled out Cepharanthine and Hypericin as the most promising for the development of antiviral agents. With this crucial information, researchers are now equipped with a better understanding and a more focused approach as they continue to battle the spread of COVID-19 and restore our world back to its normal state.
Conclusion: In Supercomputers We Trust?
As our search for the best drug combinations and vaccines to end the Coronavirus pandemic wear on, AI, deep learning, and machine learning processes continue to wow us. Their utility and ability to provide us with fact-based information about how to keep our lives safe is indispensable. When we emerge from the other side of this crisis, and we surely will, you can be certain that technology will have played an enormous role in our recovery.