Innovative AI-Driven Development of Broad-Spectrum Coronavirus Antiviral Drugs

Scientists at Harvard University's Wyss Institute have made significant advances in developing a universal antiviral treatment against coronaviruses using AI-enabled dynamic modeling. Coronaviruses are responsible for about 30% of respiratory infections worldwide, which can lead to severe illnesses and have caused global outbreaks like COVID-19. Although vaccines exist, disparities in access and emerging variants resistant to current vaccines highlight the urgent need for broad-spectrum antiviral drugs capable of oral administration and rapid distribution.

The interdisciplinary team, led by Dr. Donald Ingber, combined computational biology, medical chemistry, and innovative modeling techniques to repurpose existing FDA-approved drugs for antiviral use. Initially, they identified bemcentinib as a promising candidate, which demonstrated potential against a range of coronaviruses. Building on this, they used AI and physics-based simulations to map the Spike protein's fusion mechanism—a key step in viral entry into host cells. This approach revealed a conserved binding pocket within the Spike protein’s S2 subunit that could be targeted by drugs.

Through virtual screening of approximately 10,000 existing medications, bemcentinib emerged as a top candidate due to its strong binding affinity and oral availability. Laboratory tests confirmed that bemcentinib could effectively inhibit SARS-CoV-2 infection by blocking virus entry into lung cells. To improve efficacy and ensure the antiviral effect stemmed from interaction with the Spike protein — rather than its original target, AXL kinase — scientists designed analogs like WYS-633 and later, WYS-694. The latter showed promising results in mice models, significantly reducing viral loads and offering a potential prophylactic treatment.

This research not only suggests a new pathway to combat current and future coronavirus outbreaks but also demonstrates the power of combining AI with structural biology. The approach could be adapted to develop drugs against other viruses that use membrane fusion for entry, including influenza, Ebola, HIV, and measles. Dr. Ingber emphasized that this integrated methodology holds valuable promise for rapid drug discovery in pandemic scenarios.

Source: https://medicalxpress.com/news/2025-06-broad-spectrum-coronavirus-drug-ai.html