Drugs Created by a DeepMind Offshoot Using AI Are Set for Human Trials

Google DeepMind’s AlphaFold has already transformed the way scientists comprehend proteins. Now, the platform’s capability to create safe and effective drugs is set to be tested.

Isomorphic Labs, the UK-based biotech spinoff of Google DeepMind, will soon commence human trials for drugs developed using its Nobel Prize–winning AI technology. “We’re preparing to enter the clinic,” Isomorphic Labs president Max Jaderberg stated on April 16 at WIRED Health in London. “It’s going to be a thrilling moment as we start clinical trials and witness the efficacy of these molecules.”

Jaderberg did not provide specifics on the timeline, but it is later than the company initially intended for starting human studies. Last year, CEO Demis Hassabis indicated that AI-designed drugs would be in clinical trials by the end of 2025.

Founded in 2021 as a spinoff from Alphabet’s AI research arm, Google DeepMind, Isomorphic Labs leverages DeepMind’s AlphaFold, a cutting-edge AI platform that predicts protein structures for drug discovery.

Proteins, composed of 20 different amino acids, are crucial for all living organisms. Long chains of amino acids link up and fold into a protein’s three-dimensional shape, which determines its function. Researchers have attempted to predict protein structures since the 1970s, but this process was labor-intensive due to the astronomically high number of potential shapes a protein chain can assume.

That shifted in 2020, when DeepMind’s Hassabis and John Jumper revealed groundbreaking results from AlphaFold 2, utilizing deep-learning techniques. A year later, the company released an open-source version of AlphaFold for universal access.

In 2024, DeepMind and Isomorphic Labs launched AlphaFold 3, further enhancing scientists’ grasp of proteins. This version expanded beyond modeling proteins in isolation to predicting significant molecules like DNA and RNA, along with their interactions with proteins.

“This is precisely what is needed for drug discovery: understanding how a small molecule will bind to a target, the strength of that binding, and what else it may interact with,” Hassabis shared with WIRED at the time.

Since its launch, the AlphaFold platform has successfully predicted the structure of nearly all 200 million proteins recognized by researchers and has been utilized by over 2 million users across 190 nations. This breakthrough led to Hassabis and Jumper receiving the Nobel Prize for Chemistry in 2024, with the Nobel committee highlighting AlphaFold’s role in various scientific advancements, such as improved understanding of antibiotic resistance and the creation of images of enzymes capable of breaking down plastics.

Earlier this year, Isomorphic Labs introduced a more powerful tool, referred to as IsoDDE, its proprietary drug-design engine. A technical paper from the company claims this platform more than doubles the accuracy of AlphaFold 3.

The startup has partnered with Eli Lilly and Novartis to collaborate on AI-driven drug discovery and is also progressing on its own “broad and promising pipeline of new medicines” in oncology and immunology, Jaderberg stated.

“The exciting aspect of the molecules we are designing is that, thanks to our enhanced understanding of their functioning, we’ve engineered them to be highly potent,” Jaderberg conveyed to the WIRED Health audience. “You can administer them at much lower doses, resulting in fewer side effects and off-target effects.”

Last year, Isomorphic appointed a chief medical officer and announced raising $600 million in its first funding round to prepare for clinical trials. Concurrently, the company has been assembling a clinical development team. Its mission is to “solve all disease.”

“It’s a bold mission,” Jaderberg remarked. “But we genuinely mean it. We say it earnestly because we believe it should be achievable.”