Canadian-founded company develops first AI-designed COVID-19 drug, starts clinical trials

For the first time ever, a drug for COVID-19 designed with artificial intelligence is headed to clinical trials.

Insilico Medicine, a Hong Kong-based AI pharmaceutical company, announced last Thursday its ISM3312 therapeutic would be starting clinical trials in China after its preclinical studies found that the drug "significantly reduces" viral load and inflammation in the lungs.

 “Generative AI is transforming every area of human development,” Alex Zhavoronkov, Insilico's Canadian founder and CEO, said in a news release. “We’re extremely happy to announce that our second small molecule therapeutic generated using generative AI is now entering human clinical trials and there is a very clear timeline to demonstrate the discovery and development cycle.”

• Climate Barometer newsletter: Sign up to keep your finger on the climate pulse

The drug works by inhibiting enzymes in the virus known as proteases by irreversibly binding to them. The particular protease it targets is called 3CLpro, the main protease found in the SARS-CoV-2 virus.

"The target is a great target. The beauty of it is that this protease that is targeted, it's very dissimilar of those that we in the human body actually have, so it's a good target," Oxford University's Dr. Peter Juni, former head of the Ontario Science Table, told CTV's Your Morning on Tuesday.

Insilico leveraged AI in order to develop and test a molecule that can bind and inhibit the protease by fitting into its pockets. This molecule would then become the basis of the drug.

"What AI is really good at is detecting patterns and then basically make use of these patterns to develop a drug or a molecule that fits into certain pockets," Juni explained. "AI doesn't care about our conventions that we have scientists. It just recognizes patterns and tries to find the best fit."

Currently, the drug treatment options for COVID-19 include prescription antivirals like Paxlovid and Evusheld, but Insilico researchers note that these drugs may be ineffective at treating future variants of the virus. The U.S. Food and Drug Administration has already pulled its authorization for Evusheld, citing its inefficacy in treating certain variants of COVID-19.

But similar proteases to the one that ISM3312 targets are found in all coronaviruses, meaning that the drug could be effective at not only targeting current and future variants of SARS-CoV-2, but other coronaviruses as well.

However, despite its potential, Juni says it's too early to tell if the drug will be a success.

"It's a good start but there's a long time to go until we know whether it's safe in humans, and whether it actually works," he said.