New York, August 5, 2026
Schrödinger, Inc. has entered into a strategic collaboration and software agreement with Bristol Myers Squibb (BMS) to deploy Bunsen, Schrödinger’s agentic AI co-scientist, across BMS’s research organization, significantly expanding the companies’ longstanding relationship in computational drug discovery. The agreement is designed to give BMS scientists broader access to advanced artificial intelligence and physics-based computational technologies, enabling researchers to explore more molecular possibilities, prioritize promising compounds with greater confidence and potentially accelerate decisions during early-stage drug discovery. BMS already uses Schrödinger’s computational platform, and the expanded collaboration will integrate Bunsen with Schrödinger’s molecular modeling capabilities and RetroSynth, its AI-driven synthesis planning platform. The initiative reflects the pharmaceutical industry’s growing adoption of agentic AI systems capable of supporting complex scientific workflows rather than simply generating predictions or analyzing isolated datasets.
Bunsen Brings Agentic AI to Molecular Drug Discovery
At the center of the collaboration is Bunsen, an AI co-scientist specifically developed for molecular discovery. Unlike general-purpose AI agents, Bunsen is optimized to operate Schrödinger’s validated physics-based computational platform. The system can interpret scientific objectives, plan computational strategies, execute sophisticated molecular discovery workflows and analyze resulting data. By combining agentic AI with physics-based molecular simulation, Schrödinger aims to make advanced computational methods accessible to a wider range of scientists while allowing research teams to systematically evaluate more hypotheses before moving candidates into experimental testing. Under the new agreement, Schrödinger will also work directly with BMS scientists to develop novel functionality within Bunsen, aligning the platform more closely with large-scale pharmaceutical research requirements. The approach could help scientists navigate broader chemical design spaces while using computational evidence to prioritize molecules that warrant laboratory investigation, supporting a more prediction-driven approach to discovery.
AI and Physics-Based Modeling Expand Chemical Exploration
The collaboration will combine Bunsen with technologies designed for large-scale chemical exploration and AI-enabled synthesis planning. One important component is Schrödinger’s RetroSynth platform, which evaluates synthetic feasibility and helps identify potential routes for producing computationally prioritized molecules. RetroSynth incorporates intelligence across billions of chemical building blocks to predict plausible synthetic pathways and optimize route selection. Connecting molecular design with synthesis planning could help researchers assess not only whether a compound has attractive predicted properties but also whether it can be efficiently produced for laboratory testing. BMS plans to deploy Bunsen at scale within its research organization, giving scientists another computational capability for navigating molecular design space and advancing potential medicines. The partnership builds on years of collaboration between the companies and highlights how major pharmaceutical organizations are increasingly integrating AI, computational chemistry and physics-based simulation into established research processes.
Advertisement
Strategic Agreement Advances Agentic AI in Pharma R&D
The expanded relationship represents an important development in the adoption of agentic artificial intelligence for pharmaceutical research and drug discovery. Traditional computational tools often require scientists to manually define and execute individual analyses, whereas agentic systems are being designed to coordinate multiple stages of a scientific workflow while interpreting results and determining subsequent computational steps. Bunsen’s integration with Schrödinger’s established platform could allow BMS researchers to evaluate broader chemical spaces and make higher-confidence decisions before experimental testing, potentially improving research efficiency. However, the platform remains a decision-support and computational discovery technology, and experimental validation will continue to be essential for determining the actual biological activity, safety and therapeutic potential of candidate molecules. For Schrödinger, the agreement also demonstrates the growing commercial role of its computational platform within large pharmaceutical research organizations. The combination of Bunsen, RetroSynth, molecular simulation and BMS’s drug discovery expertise illustrates a broader industry shift toward integrated AI-enabled research environments that connect molecular design, computational prediction and synthesis planning. As pharmaceutical companies seek to improve R&D productivity, collaborations combining specialized AI with validated scientific modeling could become increasingly important in identifying and advancing innovative drug candidates.
Source: Schrödinger, Bristol Myers Squibb press release
