Engineering T cell receptors in TCR-T cell therapy using Machine Learning
KEYWORDS: Cancer | Machine Learning & AI | Immunology | T cells
Challenge
Adoptive cell transfer (ACT) is a promising approach in cancer immunotherapy. It consists in providing a cancer patient with a large number of immune cells that were previously grown and multiplied in a lab. One of the techniques in ACT is T cell receptor–engineered T (TCR-T) cell therapy. TCR-T cells are genetically engineered cells that express a receptor directed against a specific neoantigen. This antigen is a protein expressed in cancer cells as a result of tumor-specific mutations in their DNA. Though the use of TCR-T cells holds great promise in improving the sensitivity and efficacy of cancer immunotherapy, the process of engineering cell therapies that target neoantigens in cancer is complex and time-consuming. To address these challenges, our customer asked us to apply machine learning to this process.
Solution
Our team developed an ML model that can predict the functionality of T cell receptors from gene expression information and identifies gene expression signatures that define the binding property of TCRs to neoantigens.
Impact
The ability to identify computationally gene expression signatures associated with binding to neoantigen enabled our customer to accelerate the therapeutic development process in their labs.
Check our other Resources on Machine Learning & AI:
