AI/ML Model Development for Protein Aggregate Classification

This contract involves the development and training of advanced machine learning models to detect, classify, and analyze subvisible protein aggregates in biosimilar drug products. The approach focuses on using morphological fingerprinting combined with neural network-based metric learning techniques, explicitly foregoing traditional image pre-processing methods to enhance model performance. The objective is to improve the accuracy and efficiency of identifying protein aggregates, which are critical quality attributes in biosimilar drug manufacturing and safety assessment. The contract is issued as a subcontract by the FDA Office of Acquisition and Grant Services under the Department of Health and Human Services, with the place of performance located in Silver Spring. It falls under the NAICS code 541715, which pertains to research and development in physical, engineering, and life sciences. The solicitation was posted on May 11, 2026, with responses due by May 26, 2026. The project represents an opportunity to contribute cutting-edge AI/ML methodologies to pharmaceutical quality control, supporting regulatory science initiatives within the healthcare sector.