Isomorphic Lands $3 Billion AI-Powered Drug Discovery Deals with Pharma Giants

In a major validation of its AI-driven drug discovery platform, Alphabet subsidiary Isomorphic has secured two blockbuster partnerships with pharmaceutical leaders Eli Lilly and Novartis. The deals, totaling nearly $3 billion in upfront and milestone payments, aim to leverage Isomorphic‘s cutting-edge AI technology to accelerate the discovery of next-generation small molecule therapies.

Founded in 2021 as a spin-out of Google‘s DeepMind, Isomorphic has quickly emerged as a pioneer in applying AI to transform the slow, expensive process of finding new medicines. At the heart of the company‘s approach is AlphaFold, a deep learning system that can predict the 3D structure of proteins with unprecedented accuracy based solely on their amino acid sequence.

Since its initial release in late 2020, AlphaFold has been hailed as a historic advance in structural biology. The system outperformed all other computational methods and reached near-experimental accuracy in the 14th Critical Assessment of Protein Structure Prediction (CASP14) challenge. By the end of 2021, DeepMind had used AlphaFold to predict structures for over 98% of the 20,000 proteins in the human proteome.

Isomorphic has built on AlphaFold‘s protein structure prediction capabilities to create a next-generation platform for therapeutic discovery and design. The company‘s AI systems can model how small molecule compounds interact with target proteins, predict binding affinities and key properties like solubility and toxicity, and guide the optimization of drug candidates.

This AI-powered approach promises to drastically accelerate the discovery of novel small molecule drugs, which make up over 75% of all FDA-approved medicines. Traditionally, identifying promising small molecule leads has been a largely trial-and-error process, requiring the synthesis and screening of huge chemical libraries. It often takes 2-3 years and tens of millions of dollars to find a drug-like molecule suitable for clinical development.

Isomorphic‘s AI platform aims to make this process far more efficient and effective. In a November 2023 Nature Biotechnology paper, company scientists demonstrated AlphaFold‘s ability to accurately predict the binding poses of known drugs to their protein targets. The AI was even able to design new small molecules with enhanced potency and selectivity against disease-relevant proteins.

These proof-of-concept studies caught the attention of major pharma players looking for an edge in small molecule discovery. In December 2023, Eli Lilly announced a collaboration with Isomorphic to discover and develop novel therapies against challenging targets, with $45 million paid upfront and up to $1.7 billion in potential milestone payments.

Just a month later in January 2024, Novartis followed suit with a similar deal focused on three undisclosed targets, providing $37.5 million upfront and up to $1.2 billion in downstream payments. Together, the partnerships provide a strong endorsement of Isomorphic‘s AI-first approach and significant near-term revenue to fund further growth.

"We are thrilled to work with Eli Lilly and Novartis, two global pharma leaders at the forefront of applying AI and data science to accelerate drug discovery," said Isomorphic CEO and Founder Demis Hassabis. "These collaborations will allow us to further validate our AI-powered platform while expanding its reach and impact to bring life-changing therapies to patients faster."

Isomorphic‘s accomplished founding and advisory teams bring deep expertise at the intersection of AI and biology. Hassabis, who also leads DeepMind, has been a key figure in advancing AI systems for protein structure prediction, genomics analysis, and scientific discovery. Isomorphic‘s Chief Scientific Officer Miles Congreve was previously a senior drug discovery leader at Sosei Heptares and AstraZeneca, while Chief Technology Officer Sergey Yackeen has led major machine learning initiatives at 23andMe and Calico Labs.

The company‘s scientific advisory board features luminaries like Jennifer Doudna, co-inventor of CRISPR gene editing and 2020 Nobel laureate in chemistry; Demis Hassabis, artificial intelligence pioneer and founder of DeepMind; Daphne Koller, co-CEO and founder of Insitro; and Sir Paul Nurse, 2001 Nobel laureate in physiology and expert in cell cycle regulation.

Since launching in 2021 with $500 million in initial funding from Alphabet, Isomorphic has grown to over 200 employees across AI engineering, biomedical research, and drug discovery functions. The company has continued to enhance AlphaFold‘s capabilities while building proprietary AI systems for small molecule design, protein-protein interaction prediction, and biological activity modeling.

Isomorphic‘s AI technology is already showing promise in tackling historically difficult drug targets like protein-protein interfaces and allosteric sites. In a recent preprint, the company used AlphaFold to design cyclic peptides that potently block the SARS-CoV-2 spike protein from binding the human ACE2 receptor, highlighting the platform‘s potential for rapid infectious disease response.

Longer-term, Isomorphic aims to combine its structure-based AI with predictive models for cell and tissue-level effects, safety pharmacology, and clinical efficacy. By unifying molecular, preclinical, and clinical data with bespoke AI systems, the company seeks to create an end-to-end platform for AI-driven drug discovery and development.

"The true value of AI in drug discovery will be realized when we can go from novel biology and disease insights to new medicines with a high probability of success in the clinic," said Isomorphic CSO Congreve. "This is our ultimate vision, and integrating diverse data modalities with powerful AI systems and drug discovery capabilities is how we will achieve it."

Isomorphic is one of several companies leading the charge in applying AI to transform small molecule drug discovery. Relay Therapeutics and Schrodinger have leveraged computational modeling and machine learning to design clinical-stage oncology drugs, while Exscientia and Recursion have struck major AI drug discovery partnerships with pharma giants Sanofi and Bayer AG, respectively.

More broadly, the convergence of AI and drug discovery has gained significant momentum and investment in recent years. A 2023 survey by the Pistoia Alliance found that 75% of pharmaceutical and life science organizations are now using or experimenting with AI, up from 44% in 2020. Market research firm Emersion Insights projects that the global AI in drug discovery market will reach $4.8 billion by 2027, growing at a 28.9% CAGR from 2020 to 2027.

Even with this progress, realizing the full potential of AI-first drug discovery will require extensive validation in the lab and clinic. While AlphaFold and related systems excel at predicting static protein structures, accurately modeling the dynamic conformational changes and downstream biological effects of small molecules remains an immense challenge. Translating AI-generated molecular insights and predictions into experimentally confirmed actives with clinical efficacy will be an ongoing priority as the field matures.

For Isomorphic, the Eli Lilly and Novartis partnerships represent a major opportunity to demonstrate the real-world impact of its technology in discovering new therapies for patients. The initial disease areas and targets pursued in these collaborations have not been disclosed, but Isomorphic‘s platform shows promise across a range of historically difficult disorders like neurodegeneration, cancer, and autoimmunity.

"We are excited to collaborate with Isomorphic to pursue promising new targets and bring potentially life-changing treatments to patients," said Novartis CEO Vasant Narasimhan. "Isomorphic‘s AI-powered platform has the potential to substantially accelerate and improve our drug discovery capabilities, building on our commitment to data science and digital technology."

As Isomorphic‘s collaborations progress and pipeline matures, the company intends to further advance the underlying science and expand the reach of its platform. Integrating AlphaFold‘s structure prediction with emerging AI systems for protein function, biomolecular interaction, and cellular pathway modeling could yield powerful predictive and generative models for therapeutic discovery. Applying these AI capabilities beyond small molecules to modalities like antibodies, cell therapies, and RNA therapeutics is another key opportunity.

Isomorphic‘s rapid rise and $3 billion biopharma alliance spree signal an inflection point for AI-native drug discovery. While significant challenges and unknowns remain, the potential of intelligent systems to understand disease biology, identify novel therapeutic approaches, and design optimized molecules is becoming increasingly clear. As AI and machine learning continue to advance and integrate with experimental research, expect a new era of faster, smarter, and more impactful drug discovery to emerge.