Under the expanded deal, Nabla Bio will receive upfront and research payments in the double-digit millions, while deploying its proprietary AI platform, known as the Joint Atomic Model (JAM), to design protein-based therapeutics for Takeda’s early-stage pipeline.
The partnership aims to develop multi-specific drugs and custom biologics targeting complex diseases that have proven difficult to treat through conventional methods.
Nabla Bio’s CEO, Surge Biswas, likened the JAM platform to ChatGPT—only instead of answering text-based questions, it designs antibodies to match specific molecular targets.
“We are basically working on whatever the most pressing problems in Takeda's discovery portfolio is at any given time, and using JAM to help unlock and unblock those,” Biswas told Reuters.
The Boston-based biotech says its system offers one of the fastest feedback loops in the industry, delivering design-to-lab-test results within three to four weeks—a fraction of traditional timelines.
The expanded partnership reflects Takeda’s strategic pivot toward faster, AI-driven drug discovery, following its recent decision to scale back cell therapy research. The company also joined a consortium with Bristol Myers Squibb and other industry players this month to train AI models on shared pharmaceutical data.
For Nabla, the deal represents a major validation of its technology. The company expects first-in-human trials for its AI-designed molecules within the next one to two years, marking a significant milestone for AI’s role in reshaping drug development.