Isomorphic Labs Wins a Global Recognition Award 2026
Dr. Sarah Chen’s hands trembled as she closed another notebook from a failed experiment. Five years. Seventeen drug candidates. Zero successes. The cancer protein she was hunting—TIM‑3—mocked her from every failed assay. Pharmaceutical giants had thrown billions at it. The target was “undruggable.” Not because the biology was wrong—TIM‑3 absolutely drove tumor resistance—but because the protein’s binding pocket was too shallow and too flexible for traditional drug molecules to grip. It was like trying to grab smoke.
Sarah represented millions of researchers facing an uncomfortable truth: roughly 80% of disease‑causing proteins in the human body can’t be targeted with current technologies because their binding sites don’t play well with our usual tools.
Then, on a Tuesday morning in May 2024, everything changed. Her colleague submitted TIM‑3’s genetic sequence to a new system called AlphaFold 3. No experimental data. No protein crystals. Just the genetic code. Seven seconds later, they had atomic‑resolution predictions of how potential drug molecules could bind to TIM‑3. Within weeks, they synthesized three candidates. Within months, they had lead compounds showing promise. The “undruggable” target wasn’t undruggable anymore.
This type of transformation is why Isomorphic Labs, a London‑based AI‑first drug discovery company, has won the 2026 Global Recognition Award for reimagining pharmaceutical development through artificial intelligence that not only accelerates existing processes but fundamentally redefines what is possible in treating human disease. Founded in 2021 by Sir Demis Hassabis—co‑founder and CEO of Google DeepMind and 2024 Nobel Prize in Chemistry laureate for AlphaFold—Isomorphic has secured over $600 million in external funding, signed multi‑target collaborations with Eli Lilly and Novartis worth nearly $3 billion in potential payments, and co‑developed AlphaFold 3, the first AI system to surpass gold‑standard physics‑based tools for biomolecular structure prediction.
Technical Innovation and Architecture
AlphaFold 2 famously cracked the 50‑year‑old protein structure prediction problem and helped earn Hassabis and DeepMind researcher John Jumper the 2024 Nobel Prize in Chemistry, legitimizing AI and computational methods at the highest level of the discipline. AlphaFold 3, co‑developed by Isomorphic Labs and Google DeepMind and launched in May 2024, goes further: it predicts the joint 3D structure and interactions of virtually all of life’s molecules—proteins, DNA, RNA, small molecules (ligands), antibodies, peptides, and ions—with unprecedented accuracy.
Under the hood, AlphaFold 3 upgrades the Evoformer deep learning architecture from AlphaFold 2 and combines it with a diffusion‑based network similar to those used in modern image generators. Starting from a “cloud” of atoms, the model iteratively refines its guesses until it converges on highly accurate molecular structures and complexes. On benchmarks like PoseBusters, it delivers around 50% higher accuracy than the best traditional methods for protein‑ligand and antibody‑protein binding, without requiring any experimental structural input—making it the first AI system to outperform established physics‑based tools for these tasks.
For drug discovery, Isomorphic has gone a step further and built the Isomorphic Labs Drug Design Engine (IsoDDE). This next‑generation platform sits on top of AlphaFold 3 and complementary in‑house AI models. IsoDDE more than doubles AlphaFold 3’s accuracy on a demanding protein‑ligand generalization benchmark, predicts binding affinities at accuracies exceeding gold‑standard physics‑based methods, and can identify novel binding pockets on proteins using only the amino‑acid sequence as input. This enables structure predictions and design insights in seconds instead of the months or years required by traditional experimental approaches, attacking the brutal economics of drug development—historically more than a decade and billions of dollars per approved drug, with high clinical failure rates., rather than the months or years required by traditional experimental approaches, address.
Researchers are already using AlphaFold‑powered approaches to design and synthesize candidates for areas such as hepatocellular carcinoma and neglected diseases. At the same time, Isomorphic and Google DeepMind support global use through tools such as the AlphaFold Server. In cases like TIM‑3, structure‑based predictions guide compound design against targets previously dismissed as undruggable, providing a concrete path from sequence to candidate molecule.
Market Strategy and Leadership
When Demis Hassabis founded Isomorphic Labs in November 2021, his vision sounded like science fiction: not to eradicate every disease at once, but to build a systematic, repeatable, AI‑powered engine that could be applied to new diseases as they emerge. His track record gave that vision instant credibility: DeepMind’s creation in 2010, its acquisition by Google in 2014, AlphaGo’s victory over the Go world champion in 2016, AlphaFold 2’s landmark 2020 breakthrough, and now the Gemini AI family competing with GPT‑style models.
The strategy was clear: turn the one‑off success of AlphaFold into a full‑fledged drug design engine, integrated end‑to‑end from target selection to optimized candidate molecules. Early on, pharmaceutical executives had good reasons to be skeptical—AI drug discovery had seen its share of unfulfilled promises. But Isomorphic arrived with something different: battle‑tested technology already in use by millions of scientists worldwide via AlphaFold, access to Alphabet‑scale compute, and a scientific culture explicit about both strengths and limitations.
In January 2024, that credibility translated into action: Eli Lilly and Novartis signed multi‑target collaborations with Isomorphic Labs focused on AI‑driven small‑molecule discovery, deals with a combined potential value of nearly $3 billion, excluding royalties. These collaborations have since expanded, with Novartis increasing the number of joint programs as initial projects showed progress, and Johnson & Johnson later joining as another major partner for AI‑enabled multi‑modality drug discovery.
The momentum continued in 2025. In March, Isomorphic announced its first external funding round: a $600 million round led by Thrive Capital, with participation from GV and follow‑on capital from Alphabet, signaling strong investor confidence in both the science and the business model. The new capital is earmarked for further development of its AI drug design engine and to move internally discovered drugs into clinical trials, shifting the company from a pure platform play to a hybrid of platform and pipeline.
Hassabis has summarized the ambition starkly: where a traditional biotech might successfully bring one or two drugs to market in its entire lifetime, Isomorphic aims to build a system that can generate dozens of high‑quality drug candidates each year. In a world where access to previously unreachable targets could unlock much of the disease “dark matter,” whoever masters this scalable engine will help define the next era of medicine.
Industry Impact and Future Vision
The top‑line numbers—$600 million raised, nearly $3 billion in big‑pharma partnerships, and order‑of‑magnitude speed and accuracy gains—only hint at the human stakes. For Ana, a mother in rural Brazil watching her daughter fade from Chagas disease, neglected by most pharmaceutical pipelines because patients are poor, the difference is life‑defining rather than incremental. For Marcus, diagnosed with a rare cancer driven by a protein no one could previously target with a drug, the emergence of tools that can see and exploit new binding pockets is the difference between “no options” and “maybe.”
Isomorphic’s technology, combined with Open AlphaFold resources, enables groups like the Drugs for Neglected Diseases Initiative (DNDi) to pursue treatments for Chagas and leishmaniasis, diseases systematically underserved by traditional market‑driven R&D. At the same time, AlphaFold’s open‑access model—empowering more than 2 million researchers in over 190 countries to freely use its predictions—has catalyzed thousands of discovery projects across academia and industry, in sharp contrast to the classic IP‑heavy approach of pharma.
Geographically, Isomorphic is evolving from a London‑centric startup into a multi‑hub organization, expanding across sites such as Lausanne and the Boston–Cambridge area to tap into global talent pools and position its own compounds for clinical development. The launch of IsoDDE in early 2026 marks a new chapter: faster, more accurate protein‑ligand modeling and pocket discovery at a fraction of the time and cost of physics‑based methods, bringing routine virtual design loops closer to the “seconds‑to‑insight” ideal.
Inside Isomorphic’s offices, teams are assembling the components of what they call a “virtual cell”: an integrated AI system that can simulate how complex biological networks respond when you perturb them with a drug. Imagine uploading a disease mechanism, running thousands of in‑silico experiments across different modalities—small molecules, antibodies, molecular glues—predicting efficacy and toxicity with atomic‑level granularity, and narrowing to the best few candidates before any wet‑lab synthesis. That is no longer a sci‑fi storyboard; it is a technical roadmap under active construction.
For investors seeking ventures where outsized financial returns align with measurable gains in global health, Isomorphic offers a blueprint: a scalable AI engine designed for high‑value commercial indications and neglected diseases alike. For pharmaceutical leaders watching competitors lock in multi‑year collaborations, the message is clear: the window to secure first‑mover advantage in AI‑native drug discovery is closing. And for patients like Ana and Marcus, the end of the word “undruggable” is not merely a scientific milestone; it marks the arrival of a future where AI‑powered discovery enables advanced therapeutics to be developed faster, cheaper, and more broadly accessible.
This award does not recognize incremental efficiency tweaks. It honors a fundamental reimagining of what becomes possible when frontier artificial intelligence is woven into the fabric of human biology: turning a 50‑year “impossible” problem into a routine input, persuading one of the world’s most conservative industries to stake billions on a new paradigm, and proving that profit and purpose can coexist in a platform that designs blockbuster drugs and neglected‑disease treatments with the same underlying engine.
