DeepMind spin-off steps up effort to use AI to create new drugs

Alphabet-owned Isomorphic Labs is ramping up its operations by poaching pharmaceutical talent and opening a new office, as the artificial intelligence drug discovery start-up moves closer to securing its first commercial deal.

The UK-registered group was spun out of its sister company DeepMind, Google’s AI unit, in November last year, to focus on using AI technology to create new drugs to treat and prevent diseases.

Isomorphic is currently in talks with major pharmaceutical companies and is expected to announce a deal in the next few months, according to two people familiar with the plans.

Its work capitalises on DeepMind’s scientific breakthrough of its AlphaFold2 technology, which can be used to predict the shape of every protein in the human body with almost perfect accuracy.

Colin Murdoch, chief business officer at DeepMind, has been put in charge of setting up Isomorphic Labs, working closely with Demis Hassabis, who is chief executive of both DeepMind and Isomorphic Labs.

“It takes about 10 years to take a drug [to market], and often most of them fail sadly, and so inspired by the work we did with AlphaFold, we took a deeper look . . . and basically built conviction that there was a real opportunity here to apply AI to reimagine drugs discovery,” said Murdoch in one of the company’s first interviews on Isomorphic.

Isomorphic’s expansion comes amid a surge in interest in start-ups promising to use AI to transform drug discovery, with funding in the UK and US in this area jumping to more than $1.6bn this year, up from $668.5mn in 2017, according to data from PitchBook.

When the AlphaFold breakthrough was announced in November 2020, DeepMind said it would try to use the technology to find treatments for Chagas disease and Leishmaniasis, two of the most deadly diseases in the world.

Murdoch said Isomorphic had not focused on a specific drug or disease. “The goal is actually to build an underlying platform which is . . . agnostic to those therapeutic areas,” he said.

There are a number of “almost AlphaFold-scale” AI advancements that the team are working on that would provide the underlying engine for the platform, he added.

To lead its work in drug discovery, Isomorphic has hired several executives and staff from both scientific and pharmaceutical backgrounds, as well as in machine learning, developing computer systems that can learn through data.

The company is also expanding beyond its headquarters in London to a second office in Lausanne, Switzerland, home to a host of leading pharma companies including Roche, Novartis and Bayer, and Isomorphic’s chief technology officer Sergei Yakneen.

Yakneen has previously worked at Amazon and Sophia Genetics, a company that uses machine learning to identify tumours and other health conditions.

Other executives at Isomorphic include Miles Congreve, chief scientific officer, who previously worked at Astex Pharmaceuticals and GSK. Several staff have joined from DeepMind, as well as from BenevolentAI, Google and AstraZeneca.

“The goal of Isomorphic is to produce drugs which we can then partner with pharma to get them out into the clinic and to people with clinical need,” Murdoch said.

He added that Isomorphic was “beginning to think about what the right commercial path is. We have an amazing leadership team up and running and making fantastic progress”.

The company said it is in discussions with “many of the world’s leading pharmaceutical companies” without providing further details. It expects to embark upon a number of partnerships as it scales.

Murdoch said Isomorphic would hire more staff next year. Its talent acquisition lead James Girling recently aimed a post on LinkedIn at tech workers fired by Twitter.

While the AI drug discovery market has experienced growing interest in recent years, investment in the sector is not immune to this year’s tech rout. Venture capitalist funding has fallen 15 per cent from $2bn last year, according to data from PitchBook.

Some in the medical sector are sceptical that AI drug discovery will fulfil its hype, pointing to the need to navigate strict regulations and integrate into dated healthcare systems.

A recent Morgan Stanley report noted that investors would need to see “solid evidence for real-world use cases for AI-enabled drug discovery”.

However, it added that this method of drug discovery could lead to an additional 50 novel therapies over the next 10 years, presenting a potential $50bn opportunity.

Isomorphic reported a £2.4mn loss for the 11 months to December 2021, according to filings from the UK’s Companies House. This includes £470,455 in receiving contract research and development services from DeepMind as the company was launching.

Additional reporting by Madhumita Murgia and Hannah Kuchler in London