USC Researchers Receive ARPA-H Grant to Develop AI Framework for CGT Outcomes in Rare Pediatric Diseases

Mar,26 2026-

Researchers at the Keck School of Medicine of USC have been awarded up to $6.8 million from the Advanced Research Projects Agency for Health (ARPA-H) to develop new computational models aimed at accelerating access to cell and gene therapies for children with rare diseases.

The two-year project, part of the UNICORN (UNIfying Cell Therapy Outcome prediction and Regulatory Navigation) program, will focus on building an AI-driven framework that links therapy design, manufacturing characteristics, and patient outcomes. The initiative is led by Mohamed Abou-el-Enein, MD, PhD, executive director of the USC/Children’s Hospital Los Angeles (CHLA) Cell Therapy Program, with support from ARPA-H Program Manager Daria Fedyukina, PhD.

Cell and gene therapies have transformed treatment options for conditions such as childhood cancers and inherited genetic disorders, but their development remains challenging. Unlike conventional drugs manufactured in large batches, these therapies often involve genetically modified patient cells produced individually, resulting in complex manufacturing processes, high costs, and relatively small clinical datasets.

Because clinical trials for rare diseases frequently involve limited patient populations, conventional statistical models may not fully capture the relationships between therapy characteristics and patient outcomes. The UNICORN project aims to address this challenge by using artificial intelligence and advanced cell analytics to identify biological patterns that predict treatment responses.

At the core of the initiative is a high-dimensional cell-analysis platform developed in Abou-el-Enein’s laboratory. The system was originally designed to study chimeric antigen receptor (CAR) T cell therapies, enabling researchers to simultaneously analyze multiple protein markers on individual cells. This approach provides detailed insights into cell phenotype, functional activity, and manufacturing dynamics, helping identify features associated with stronger therapeutic potency.

With ARPA-H support, the research team will expand the platform to analyze a broader range of cell and gene therapy products, including:

• CAR T cell therapies for hematologic malignancies
• Hematopoietic stem cell (HSC)–derived gene therapies
• Gene-edited cellular therapies for pediatric genetic diseases

The project will integrate therapy manufacturing data, product characterization metrics, and longitudinal patient samples collected across multiple pediatric disease programs. These datasets will be used to train machine learning models capable of linking product attributes with clinical outcomes, creating a predictive framework for therapy development.

A key goal of the UNICORN program is to develop a regulatory decision-support tool that helps interpret evidence for advanced therapies when traditional large-scale clinical trials are not feasible. Such tools could help regulators and clinicians evaluate the quality and expected performance of individualized therapies, potentially enabling faster access to life-saving treatments for children with rare diseases.

To support multi-center collaboration and data integration, the project will utilize Bluecord, an electronic quality and data management platform that enables standardized sample tracking, secure data sharing, and structured linkage of clinical and manufacturing datasets. The system was implemented at USC with prior support from the California Institute for Regenerative Medicine (CIRM).

Researchers envision UNICORN as a “learning system” for rare disease therapies, where each treated patient contributes new data that improves predictive models and informs future treatments.

If successful, the project could provide a new blueprint for developing data-driven frameworks that support the clinical development, evaluation, and regulatory navigation of next-generation cell and gene therapies.