SFU Chemists Use AI to Speed Antiviral Drug Discovery

Simon Fraser University researchers have developed a method that dramatically accelerates how quickly scientists can build and screen antiviral drug candidates, technology aimed squarely at a genuine gap in modern medicine: humanity has an abundance of painkillers and antibiotics, but a genuinely thin arsenal of effective antivirals, according to the study published in the journal Science and covered in BC Technology's reporting on the research.

Chemistry professor Robert Britton, the study's lead author, explained the underlying problem directly: "We have lots of pain killers and a wide collection of antibiotics, but we don't have a good of a panel of antivirals, which is why viral outbreaks like COVID-19, or Ebola, or hantavirus scare people so much. Finding viable drug candidates is extremely challenging." Traditional drug discovery involves screening large libraries of molecules to identify promising "hits," an approach that helped companies including Merck and Gilead Sciences develop early COVID-19 treatments, but generating those large molecule libraries has historically been limited by complex, slow chemistry.

Britton's team, which included scientists at drug company Merck, began with a single, scalable molecular building block, one versatile starting point that can be produced in large quantities. Using a light-driven chemical reaction, the researchers attached different nucleobases to this core structure, rapidly generating a library of more than 70 nucleoside analogs, compounds that mimic the building blocks of DNA and RNA and are widely used to treat cancer and viral infections including HIV and hepatitis.

Britton described the speed advantage in concrete terms: "This is a game changer for making and modifying nucleosides. In an emerging outbreak, the more compounds you can screen, the better your chances of finding something effective. With this method, we can produce libraries 10 to 100 times larger in just weeks, rather than months or years."

This wasn't purely theoretical chemistry. To test whether the approach could actually find useful drug candidates, researchers screened the resulting library against HIV at SFU's Pantophlet Laboratory, led by health sciences professor Ralph Pantophlet. Three compounds showed activity comparable to approved HIV therapies already on the market, a genuinely meaningful validation result for a method still in its early stages.

The Antiviral Discovery Method by the Numbers

Notably, Britton emphasized that most of the compounds his library produced were entirely new to science. "Most of the compounds in our library were entirely new," Britton said. "A few had been made before, but it took other groups longer to synthesize those molecules, and they were not able to modify and improve them as readily."

This breakthrough builds directly on earlier work from the Britton Lab at SFU. Traditionally, synthesizing nucleoside analogs required a complex, roughly 20-step process that could take months or even years to complete. In 2020, Britton and collaborators reduced that process to just a few steps completed in about a week, a foundation this new library-generation method builds on and accelerates even further. This progression connects to a broader pattern of AI and computational chemistry breakthroughs we've tracked at SFU specifically, and to the wider push toward AI-accelerated drug discovery we've covered in our coverage of the growing use of AI across pharmaceutical research more broadly.

This research is worth understanding for any business in pharmaceuticals, biotechnology, or public health preparedness, given the persistent, structural gap in effective antiviral treatments that Britton's own comments highlight directly. A method capable of generating screenable drug candidate libraries 10 to 100 times faster than traditional approaches represents genuinely significant infrastructure for responding to future viral outbreaks, where speed of candidate generation can directly affect how quickly effective treatments reach patients.

For pharmaceutical and biotech companies evaluating research partnerships or licensing opportunities, SFU's Britton Lab and its collaboration history with major drug companies like Merck signals a genuinely productive academic-industry pipeline worth monitoring for future antiviral therapeutic candidates emerging from this specific research program.

What did SFU researchers actually develop?
SFU chemists developed a method using a light-driven chemical reaction to rapidly generate large libraries of nucleoside analogs, compounds used to treat viral infections and cancer, producing more than 70 candidates and identifying three with activity comparable to approved HIV therapies.

How much faster is this new antiviral discovery method?
The method can produce screenable drug candidate libraries 10 to 100 times larger in just weeks, compared to the months or years traditional nucleoside analog synthesis has historically required.

Has this research been tested against a real disease?
Yes. Researchers screened the resulting compound library against HIV at SFU's Pantophlet Laboratory, finding three compounds with activity comparable to already-approved HIV therapies.

Simon Fraser University chemists developed a light-driven chemical method that rapidly generates large libraries of antiviral drug candidates, producing more than 70 nucleoside analogs and identifying three compounds with HIV activity comparable to approved therapies. The technique, developed with collaborators at Merck, can produce screenable compound libraries 10 to 100 times larger in weeks rather than the months or years traditional methods require. The research builds on a decade of work at SFU's Britton Lab, which previously reduced a complex 20-step synthesis process down to about a week in 2020.