December 3, 2022

Advanced AI drives antiviral medication improvement

But Beijing's unrelenting zero-Covid approach to outbreaks in multiple cities this year has clogged supply chains and locked down tens of millions of people


But Beijing’s unrelenting zero-Covid approach to outbreaks in multiple cities this year has clogged supply chains and locked down tens of millions of people – Copyright AFP SAEED KHAN

A team of scientists who work under the non-profit, open-science COVID-19 Moonshot project, have been awarded $68,662,387 from the U.S. National Institutes of Health. The funding will be used to discover and develop globally accessible and affordable novel oral antivirals.

Such medication will help to combat the current COVID-19 global pandemic and, very likely, future pandemics.

One of the scientists involved, Ben Perry states: “If we had clinic-ready antivirals suitable for SARS-CoV-2 when the pandemic struck in late 2019, we could have perhaps saved millions of lives.”

While the past cannot be underdone, the future can be better. Perry continues: “The world needs a diverse stockpile of novel antiviral compounds that can be quickly advanced for the current pandemic and when the next pandemic strikes, and it is essential that these be affordable and equitably accessible to everyone.”

To develop new antivirals, the scientists are using artificial intelligence. This is a bespoke solution termed the Artificial intelligence (AI)-driven Structure-enabled Antiviral Platform (ASAP). The platform is based on advanced structural biology, machine learning, and computational chemistry.

The technology is hosted on [email protected], the world’s largest distributed computing platform. The intended output will be a robust antiviral discovery pipeline.

The funding is expected to help the project exist for three years, with more funding to follow. Over the initial period, the scientists aim to produce preclinical candidates against multiple viral targets.

The most promising of these will be transitioned into multiple drug candidates ready for evaluation in humans in the event of an ongoing or emerging pandemic threat.

The COVID Moonshot project that began in March 2020 and this initiative has identified potent antivirals targeting the main protease of the SARS-CoV-2 virus. These are undergoing a preclinical development program funded by the Wellcome/COVID-19 Therapeutics Accelerator.

The progress of Moonshot has demonstrates the power of AI-driven drug design. Here algorithms generate molecules with optimized properties that can quickly be made and tested in the laboratory.

ASAP will target viral families like coronaviruses (responsible for the current COVID-19 pandemic as well as earlier SARS and MERS epidemics). The project will also address flaviviruses, responsible for large endemic diseases such as dengue and Zika.