Months into the COVID-19 pandemic, supercomputers crunching the coronavirus is the norm, not the exception. Most of these systems are focused on various elements of a similar approach to vaccine or therapeutic discovery: screening compounds to see how strongly – if at all – they bond to the virus’ spike protein or main protease. Researchers are hoping to condense this arduous process of drug development, which can often take many years or even decades, into less than two years’ time using the brute force power of the world’s most powerful computers. Now, researchers from Penn State University are throwing their hats into the ring – with a twist.
Enter: quantum machine learning, a burgeoning crossover field that combines machine learning with quantum information processing.
“Discovering any new drug that can cure a disease is like finding a needle in a haystack,” said Swaroop Ghosh, professor of electrical engineering and computer science and engineering at Penn State, in an interview with the university. Ghosh and his doctoral students (Mahabubul Alam, Abdullah Ash Saki, Junde Li and Ling Qiu) were tackling a challenging element of the COVID-19 drug discovery formula: identifying the billions upon billions of molecules for the supercomputers to screen.
“High-performance computing such as supercomputers and artificial intelligence can help accelerate this process by screening billions of chemical compounds quickly to find relevant drug candidates,” Ghosh said. “This approach works when enough chemical compounds are available in the pipeline, but unfortunately this is not true for COVID-19. This project will explore quantum machine learning to unlock new capabilities in drug discovery by generating complex compounds quickly.”
This isn’t entirely new ground for the researchers, who previously worked to develop a toolkit for using quantum computing to solve combinatorial optimization problems, which aim to find the ideal item from a large – and even unsearchable – set of items. Drug discovery falls under the combinatorial optimization umbrella, which made the transition to COVID-19 drug discovery relatively painless for the team.
“Artificial intelligence for drug discovery is a very new area,” Ghosh said. “The biggest challenge is finding an unknown solution to the problem by using technologies that are still evolving — that is, quantum computing and quantum machine learning. We are excited about the prospects of quantum computing in addressing a current critical issue and contributing our bit in resolving this grave challenge.”
The researchers hope that their quantum machine learning approach will be both faster and cheaper than current approaches to generating those complex compounds.