USC team develops new AI model to advance rare disease therapies
UNICORN project uses AI to analyze therapies and improve outcomes
Written by |
Researchers at the Keck School of Medicine, University of Southern California, are receiving up to $6.8 million in funding to develop an artificial intelligence (AI)-based computational model to help advance gene and cell therapies for children with rare diseases, including conditions such as aromatic L-amino acid decarboxylase (AADC) deficiency.
Funded by the Advanced Research Projects Agency for Health (ARPA-H), the two-year project, called “UNICORN: UNIfying Cell Therapy Outcome prediction and Regulatory Navigation,” aims to understand how specific biological characteristics of therapies are linked to patient outcomes.
AI project aims to improve how rare disease therapies are developed
“This project reimagines how we develop therapies for rare diseases,” Mohamed Abou-el-Enein, MD, PhD, principal investigator of the new project and executive director of the USC/Children’s Hospital Los Angeles Cell Therapy Program, said in a university news story. “By working closely with colleagues across the field, we are building a scientific foundation that helps translate complex biological data into clearer evidence and guides the creation of more effective therapies.”
While gene and cell therapies have advanced treatment for genetic diseases in children, they often rely on living, patient-specific cells, making the manufacturing process both complex and expensive. Because these treatments are often tailored to one individual at a time, clinical development is also limited by the small patient populations typical of rare diseases.
The UNICORN project will use advanced cell analysis technology and machine learning tools to identify biological patterns and therapy features associated with treatment responses. This approach may help guide how therapy-related evidence is interpreted when conventional measures are difficult to establish, potentially enabling patients to access new treatments sooner.
At the heart of the project is an advanced cell-analysis platform, initially designed to study and improve chimeric antigen receptor (CAR) T-cell therapies, a type of cell therapy used to treat certain blood cancers.
Platform will expand beyond CAR T-cell therapies in new project
In this new project, researchers will expand the platform’s use to analyze other cell and gene therapy products.
“The technology driving this work was developed here in our lab at USC. Now it’s part of a unique effort that combines advanced cell analytics with artificial intelligence and machine learning workflows to create a blueprint for developing small-batch therapies that others can learn from and build on,” Abou-el-Enein said.
To develop the model, researchers will collect detailed data on cell and gene therapies to identify features that may affect their quality and consistency during manufacturing. They will also gather patient data and samples at multiple time points, across a range of pediatric diseases, in collaboration with academic centers in the U.S.
Patient samples will be analyzed using the cell-analysis platform, and the resulting data will be used to train the AI model to identify patterns linked to patient outcomes.
Model will analyze multiple types of gene and cell therapies
The model will be used to analyze multiple therapies, including CAR T-cell therapies, hematopoietic stem cell (HSC)-derived therapies, and gene-editing products that repair genetic defects in patients’ cells. CAR T-cells are immune T-cells, from a patient or donor, that are modified in the lab to contain a receptor that recognizes a specific protein on target cells. HSC-derived therapies genetically modify hematopoietic stem cells — the progenitors of all blood cell lineages — to correct underlying genetic defects at their source.
“Every child treated adds new data that strengthens the model for the next one. It’s a living, learning system designed to get smarter with each patient,” Abou-el-Enein said. “Our vision is simple: when a child’s life depends on one therapy, we should be able to move forward with confidence because we’ve built the systems and the evidence to guide us.”
The project is also supported by Bluecord, an electronic quality and data management platform implemented at USC and supported in part by a prior $2 million grant from the California Institute for Regenerative Medicine.
The project was featured in Nature Medicine in a correspondence outlining the scientific vision behind the UNICORN framework.
