Isomorphic Labs, the London-based spin-off of Google AI R&D division DeepMind, focused on medicine discovery, promptly announced strategic partnerships with two major pharmaceutical giants, Eli Lilly and Novartis, to leverage AI in discovering new drugs for treating illnesses.
The total value of these collaborations amounts to approximately \(3 billion. Eli Lilly will provide Isomorphic with an initial payment of \)45 million and up to \(1.7 billion linked to performance targets, excluding profits. On the other hand, Novartis will offer an upfront payment of \)37.5 million along with covering specific research expenses and potential performance-based incentives of up to $1.2 billion (excluding royalties) over time.
In a press statement, Demis Hassabis, the co-founder of DeepMind and CEO of Isomorphic, expressed, “We are excited to establish these partnerships and leverage our exceptional technology platform. These alliances are particularly compelling due to our shared commitment to advancing innovative drug design methodologies and embracing cutting-edge scientific advancements.”
Fiona Marshall, the chair of medical studies at Novartis, highlighted the transformative potential of advanced AI systems in revolutionizing the drug discovery process, stating, “Cutting-edge AI systems have the potential to revolutionize our approach to discovering new medications and enhance our capacity to provide life-altering treatments to patients.” This partnership unlocks novel avenues for AI-driven drug discovery by harnessing the unique strengths of our organizations, spanning AI, data science, medicinal science, and deep disease domain expertise.
Introduced in 2021, Isomorphic, a subsidiary of Alphabet’s DeepMind, leverages AlphaFold 2 AI technology to predict the structure of human brain proteins. The objective is to identify these structures to uncover new target pathways for developing disease treatments.
While not without flaws, recent research published in the journal Nature highlighted that AlphaFold may exhibit occasional errors and is often more effective as a “hypothesis generator” rather than a substitute for empirical data. Nonetheless, the model can generate highly accurate protein predictions on a larger scale compared to previous methodologies.
Researchers recently utilized AlphaFold to design and synthesize Hepatocellular cancer, the most common form of primary liver cancer. Furthermore, DeepMind collaborates with the Geneva-based Drugs for Neglected Diseases program, a voluntary pharmaceutical organization, to utilize AlphaFold in creating therapies for Leishmaniasis and Chagas disease, two of the most fatal diseases in developing regions.
In October, DeepMind announced that AlphaFold’s latest version can provide structural estimates for nearly all molecules in the Protein Data Bank, the world’s largest open-access repository of biological molecules. The model can accurately predict the structure of ligands, nucleic acids, and post-translational modifications, essential for understanding how cells communicate and genetic information.
Isomorphic now employs the newly developed AlphaFold model, a collaborative effort with DeepMind, to design therapeutic drugs and aid in characterizing vital molecular structures crucial for disease management.
Under pressure to achieve profitability, Isomorphic incurred a loss of £2.4 million ($3 million) in 2021 as it expanded its workforce in preparation for launching its second office in Lausanne, Switzerland.
