Bioinformatics Analysis of T-cell Receptor Using Structural and Sequencing Data

T-cell receptor (TCR) specificity and the ability to recognize multiple pathogens (cross-reactivity) are poorly understood facets of the immune response that have significant implications in medicine. Advances in disease diagnosis, vaccine schedules, and T-cell based immunotherapies could be greatly affected if TCR specificity could be predicted. The key components of TCR specificity reside in the complementary determining region (CDR) loops in both the alpha and beta chains of the TCR as amino-acid residues in these regions directly interact with the peptide loaded major histocompatibility complex (pMHC). The first step of this research begins by exploring the TCR/pMHC landscape using X-ray crystallography structures. Here the relative usage of the alpha and beta chains can be quantified by identifying residues within them that contact the pMHC. From this analysis, it was found that any given TCR may have equal or favored use of the alpha and beta chains. This is a key piece of information with regard to predicting TCR specificity as the identity of residues in favored chains become more useful for determining a TCR’s specificity. Due to a deficit of TCR/pMHC X-ray crystallography structures, TCR sequence data is the best candidate for developing a predictive method for TCR specificity. Previous methods for TCR specificity classify TCRs based on an assumption of equal TCR alpha/beta usage, from our preliminary analysis we know TCR alpha/beta usage is not always equal. Accounting for this variability in usage, we introduce a new method that outperforms previous methods for predicting TCR specificity