Innovative AI Tools Propel Breakthroughs in Coronavirus Drug Discovery

In a groundbreaking development, researchers at the Wyss Institute for Biologically Inspired Engineering at Harvard University have successfully leveraged artificial intelligence (AI) and animation techniques traditionally used in film to facilitate rapid drug discovery against coronaviruses. This initiative, which gained momentum at the onset of the COVID-19 pandemic, aims to address the urgent need for effective antiviral treatments as the world continues to grapple with the implications of coronavirus infections.

According to the World Health Organization (WHO), approximately 30% of all respiratory tract infections are attributed to coronaviruses, posing significant health risks that can escalate into epidemic and pandemic scenarios, as starkly illustrated by the COVID-19 outbreak. Despite advancements in vaccine development, accessibility remains a critical issue, particularly in low-resource countries, necessitating innovative antiviral solutions that can be rapidly deployed.

The research team, comprising multidisciplinary experts in computational biology and drug development, was led by Dr. Donald Ingber, M.D., Ph.D., the Founding Director of the Wyss Institute. With initial support from the Defense Advanced Research Projects Agency (DARPA), the team aimed to repurpose existing FDA-approved drugs to combat COVID-19. Their findings have been documented in the 2025 article published in *Frontiers in Molecular Biosciences*.

As the team embarked on their drug discovery journey, they hypothesized that targeting hidden regions of the Spike protein—rather than the more mutable external sites—could yield more robust antiviral agents. "We thought that constant regions that remain hidden while the virus initially binds to its host cell could be ideal sites for drug targeting," stated Dr. Charles Reilly, the first author of the study and a former Principal Scientist at the Wyss Institute.

Utilizing advanced AI-driven molecular modeling techniques, the researchers were able to create dynamic simulations of the Spike protein's movements during the viral entry process. These simulations facilitated the identification of a binding pocket that could potentially inhibit the virus's fusion with host cells. Following an extensive screening of around 10,000 existing drugs, the team identified bemcentinib—a drug initially approved for cancer treatment—as a viable candidate with antiviral properties.

Further optimization of bemcentinib led to the development of an analog named WYS-633, which demonstrated broad-spectrum antiviral activity against various coronaviruses, including SARS-CoV-2. The team's iterative approach, which combined AI methods with experimental validations, ultimately yielded WYS-694, a compound reported to be 12.5 times more potent than its predecessors, significantly reducing viral load in infected models.

Dr. Ingber emphasized the significance of this research: "By aiming for an orally available drug that broadly inhibits multiple coronaviruses, we deliberately set the bar at maximum height. Our integrated approach merging AI-driven computational and experimental technologies has proven to be incredibly powerful in achieving this goal."

The implications of this research extend beyond the immediate challenges posed by COVID-19. The innovative methodology developed through this project not only holds promise for addressing future respiratory pandemics but also suggests potential applications for other viral families that utilize membrane fusion proteins, including influenza and HIV.

This collaborative effort underscores the importance of interdisciplinary research in navigating complex health challenges. As the world continues to confront the evolving landscape of infectious diseases, the marriage of technology and biology may provide the keys to unlocking more effective and accessible treatment options in the future.