T-cell therapy is an immunotherapy that involves engineering a patient’s T-cell receptors, a type of white blood cell, to allow for the T-cells to bind with pathogenic and diseased cells and trigger programmed cell death. The major difficulty in implementing T-cell therapy in practice is determining the specificity of what the T-cell’s receptors should look like. This project provides a novel method to accurately predict parts of a T-cell receptor given an diseased cells’ receptor. This is done by using deep learning, a subset of artificial intelligence, to predict the CDR3 protein sequence, variable segment and joining segment of the T-cell. The deep learning models were able to reach an accuracy of 90%, and can effectively determine candidate T-cells for T-cell therapy. By doing so, this project can be leveraged to increase the response time and effectiveness of T-cell therapy in practice.
Vansh Sethi
T-cell therapy is an immunotherapy that involves engineering a patient’s T-cell receptors, a type of white blood cell, to allow for the T-cells to bind with pathogenic and diseased cells and trigger programmed cell death. The major difficulty in implementing T-cell therapy in practice is determining the specificity of what the T-cell’s receptors should look like. This project provides a novel method to accurately predict parts of a T-cell receptor given an diseased cells’ receptor. This is done by using deep learning, a subset of artificial intelligence, to predict the CDR3 protein sequence, variable segment and joining segment of the T-cell. The deep learning models were able to reach an accuracy of 90%, and can effectively determine candidate T-cells for T-cell therapy. By doing so, this project can be leveraged to increase the response time and effectiveness of T-cell therapy in practice.