Deep unsupervised learning methods for the identification and characterization of TCR specificity to Sars-Cov-2

Abstract

The T-cell receptor (TCR) is one of the key players in the immune response to the Sars-Cov-2 virus. In this study, we used deep unsu-pervised learning methods to identify and characterize TCR speci-ficity. Our research focused on developing and applying state-of-the-art modelling techniques, including AutoEncoders, Variational Au-to Encoders and transfer learning with Transformers, to analyze TCR data. Through our experiments and analyses, we have achieved promis-ing results in identifying TCR patterns and understanding TCR speci-ficity for Sars-Cov-2. The insights gained from our research provide valuable tools and knowledge for interpreting the immunological re-sponse to the virus, ultimately contributing to the development of effective vaccines and treatments against the viral infection.