T-cell receptor alpha to beta chains binding prediction

Binding of T-cell receptors (TCRs) and their cognate peptide-major histocompatibility complex (pMHC) target is determined by both TCR(Formula presented) and TCR(Formula presented) chains. However, not all TCR(Formula presented) and TCR(Formula presented) can bind to each other. Predicting their pairing is crucial for understanding the TCR–pMHC interaction and developing effective de novo TCRs. Here, we show that in the general TCR repertoire, TCR(Formula presented) and TCR(Formula presented) chain compositions are independent. However, in pMHC-binding TCRs, clear associations between TCR(Formula presented) and TCR(Formula presented) chains are found, also for TCRs binding to the same pMHC. The association between the CDR3 amino acid composition and (Formula presented), (Formula presented) usage of TCR(Formula presented) and TCR(Formula presented) reveals distinct binding patterns between specific (Formula presented) and (Formula presented) genes, as well as negative correlations between the charge and polarity of the TCR(Formula presented) and TCR(Formula presented) chains, but positive associations between their molecular weights. These associations are used for the development of a prediction model for TCR(Formula presented) and TCR(Formula presented) pairing. We present here TCR-BARN (TCR Beta-Alpha chains paiRing using Nlp) that employs an initial embedding for each amino acid in the TCR alpha and beta CDR3 sequences, followed by long short-term memory (LSTM) networks to capture sequence dependencies. The (Formula presented) and (Formula presented) genes are represented using one-hot encoding. LSTM outputs are concatenated and passed through a fully connected feedforward layer for binding prediction. TCR–BARN reaches an area under the curve (Formula presented) for epitope-bound TCRs. TCR–BARN can be used for generating cognate TCRs resembling natural TCRs and evaluating the generated TCR quality.