AI designed drugs in trials this year, says Google DeepMind chief

New drugs designed by artificial intelligence  are set to be in clinical trials by the end of this year.

Speaking during a panel session at the World Economic Forum in Davos, Demis Hassabis, chief executive of Google DeepMind said that its spin-out company Isomorphic Labs would have drugs in clinical trials this year.

Last year Hassabis and John Jumper, a senior research scientist at Google DeepMind shared the Nobel Prize for Chemistry for their work on AlphaFold, an AI system which is able to predict protein structures.

“AI applied to science is a lot richer than just the language models and things like AlphaFold, these are kind of bespoke models that are built using the same principles as the general models but are then applied to specific to particular domains,” he told the audience in Davos. “We and others are working on trying to design drugs with AI and with our spin-out company Isomorphic I think we will hopefully have some AI-designed drugs in the clinic trials by the end of the year - that's the plan.”

Separately Hassabis told the Financial Times that the company was looking at oncology, cardiovascular and neurodegeneration. Some AI-developed drugs – from companies including Isilico Medicine - are already in development or trials.

Hassabis said that AI was a good tool for exploring biology and could make it quicker to develop new treatments. “I've always felt you've needed a tool like AI to understand a lot of phenomena in biology whether that's a virtual cell or the brain because it's so complex, it's difficult to describe it with just mathematical equations,” he said.

AlphaFold 2 has helped predict the structure of virtually all the 200 million proteins that researchers have identified, something that would have otherwise taken a billion years of PhD time to complete at the rate of five years per protein, Hassabis said. But he also said that, for drug discovery, this is just one piece of the puzzle - and that AI could eventually help to bring down time to get from target to a candidate to months or weeks, for example by helping to design compounds that bind to the right part of a target protein's surface and nothing else.

“Eventually you could imagine personalised medicine where it's optimised maybe overnight by an AI system for your personal metabolism to be perfect for you,” he said.

Hassabis said that with AlphaFold 3 the intention is moving up the “interaction stack”.

“AlphaFold 2 has essentially cracked the picture of a static protein, but biology is not static; all the interesting things are when there is dynamics and interactions happening and AlphaFold 3 is the next step in that it has pair-wise interactions a protein interacting with another protein or a protein interacting with a ligand or a protein interacting with RNA/DNA. Eventually my dream would be to simulate a virtual cell,” he said.

Speaking on the same panel neuroscientist Ardem Patapoutian said AlphaFold has made a huge impact by revolutionising both basic understanding of biology as well as drug discovery, adding: "It's probably just the tip of the iceberg of how it could help the drug discovery enterprise.”

Earlier this month the US Food and Drug Administration issued draft guidance to provide recommendations on the use of AI intended to support a regulatory decision about a drug or biological product’s safety, effectiveness or quality.

More stories about pharma and AI

• Life sciences and pharma priorities in 2025: M&A, AI and China
• The FDA approved these 50 new drugs in 2024
• 34 emerging technologies that could change the future
• Google DeepMind duo share Nobel Prize for Chemistry with biochemist