The CEO of Insilico Medicine, which recently became the first company to launch Phase II trials for a drug developed using generative AI, spoke with BioSpace about the rise of the technology.
Pictured: Abstract connected lines and dots/iStock, your_photo
Artificial intelligence has a growing presence in biopharma, with applications beyond the discovery apparatus. It was a hot-button issue in San Francisco last week as the industry gathered around the annual JP Morgan Healthcare Conference. Eli Lilly and Novartis even announced multi-million-dollar discovery deals with Alphabet’s Isomorphic Labs as the week of events was just kicking off.
In the midst of all the AI buzz, BioSpace sat down with Alex Zhavoronkov, CEO of Insilico Medicine, which last summer became the first firm to enter Phase II clinical trials with a therapy developed using generative AI. Zhavoronkov spoke about the technology’s role in the industry and when Insilico may have a product on the market. In addition to its Phase II trials, which will test a therapy for the lung disease idiopathic pulmonary fibrosis in the US and China this summer, Insilico just this month signed a licensing agreement with the Menarini Group to commercialize another of its drugs.
A Cautious Tone
While Zhavoronkov agreed with the waves that AI is making and will continue to make in biopharma, he encouraged the industry to think about how it can move beyond being an investment buzzword and actually be put to good use. He noted that the shine of the term “AI” is beginning to wear off, and smaller companies really need to present some breakthrough in AI drug discovery to get funding, as many pharmaceutical companies are acquiring AI-focused companies or unlocking the skills independently.
He also acknowledged that while AI is a powerful tool, pharmaceuticals can be a slower process than other AI applications. “Without the power of experiment, you cannot validate whether the algorithm works,” he said, “but the experiment might take you six months for something very basic, and it might take you a few years if you want to get into humans.”
Competition from Asia
Zhavoronkov said that for Insilico, more competition is cropping up in mainland China. For example, a recent report from Forbes went into detail on ByteDance, the parent company of social media platform TikTok, which is getting into the game of AI drug discovery and recruiting talent from the U.S. But Zhavoronkov isn’t concerned at all.
“I know that it’s going to fail,” he said of ByteDance’s AI drug discovery Endeavor. “I’m sorry, that’s my prediction. . . . Usually, for those tech giants, when you have substantial funding and no objective function to deliver a drug in the market, they always fail.”
Other Chinese companies, such as Tencent, have ventured into AI drug discovery but have not produced anything substantive that would cause Insilico to worry about competition, he added.
Projected Timeline for Market Entry
As for when the industry will see an AI-discovered drug, especially one of Insilico’s, hit the market, Zhavoronkov was a bit evasive. He said the company has a set timeline for getting one of its drugs to market and will “live by that timeline,” but would not disclose what that timeline is to BioSpace. Instead, he simply said, “I wouldn’t be surprised if [in] the next five years we would get approval, ourselves or with a partner.”
Zhavoronkov said he wants to position Insilico as a “beacon of light” for other companies to follow in terms of leveraging AI to get a drug to market. He argued that the “real power” of AI is in drug repurposing. He also said AI applies to biologics because that sector’s aspects are centered around language.
“It’s much easier for large language models and multimodal transformers to work with protein sequencing data, with amino acid sequences and DNA because it’s language,” he said, “and it’s something that a lot of people are doing.”
Tyler Patchen is a staff writer at BioSpace. You can reach him at tyler.patchen@biospace.com. Follow him on LinkedIn.
