SBIR Phase I: AI-Directed Automation for Accelerated Bioformulation of Antibody Therapeutics

About This Grant

The broader impact /commercial potential of this Small Business Innovation Research (SBIR) Phase I project is to potentially turn hours-long hospital infusions of antibody medicines into self-administered at home injections. By pairing laboratory robotics with artificial intelligence, the project seeks to cut formulation time and material use by up to half potentially lowering the barriers of bringing new antibody medicines to patients and expanding access to self-administered therapies for cancer and autoimmune diseases. This project will serve biopharmaceutical firms developing antibody drugs, an addressable segment estimated at over two hundred million dollars annually. The technology offers a durable advantage through continuously improving artificial intelligence, while revenue will be generated via fee-for-service campaigns that scale efficiently with automation. First adopters are expected to be midsized-large biopharmaceutical companies seeking rapid, low-cost formulation. The technology’s commercial launch is planned within two years, with projected annual platform revenues of approximately three to five million dollars by the third year of production. Broadly, the work advances national health and prosperity by enabling and accelerating the delivery of next-generation biologic medicines. This Small Business Innovation Research (SBIR) Phase I project will utilize a platform discovery pipeline using artificial intelligence to model and predict antibody formulation behavior with Generally Recognized as Safe (GRAS) excipients. Specifically, this project plans to use an artificial intelligence-driven search to formulate three therapeutic monoclonal antibody drugs that are candidates for subcutaneous administration. This project seeks to rapidly identify formulations enabling therapeutic dosing ([C] > 100 mg/mL), injectable viscosity levels (η < 30 cP), and colloidal stability (Monomer > 98%). To accomplish this, the projects aims to build and implement a platform discovery pipeline using artificial intelligence / machine learning to model and predict formulation behavior with GRAS excipients. These models aim to unlock (1) discovery campaign efficiency gains from a low number of experiments / data points, (2) Explainable artificial intelligence models to quantitatively map formulation-function behavior, and (3) predictive tools to optimize and enable classically challenging drugs for at home pre-filled autoinjectors. Successful and efficient Phase I outcomes would immediately enable advanced bioformulation for critically important medicines. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.