The critical step of T cell-mediated immunity is the interaction between T-cell receptor (TCR) and epitope peptide-major histocompatibility complex (pMHC) molecules. The ability to use TCR sequences to infer antigen-specificity could greatly aid potential diagnostics and therapeutics. Here, we build a user-friendly web service named DeepTR for T cell response predicion by modeling TCR binding specificities of the neoantigens. DeepTR implements comprehensive featurization strategies in a biologically meaningful way followed with multimodal deep learning network to numerically embed sequences of MHC, antigens and six loops of TCR (α- and β-chain of CDR1-3) for predicting the pMHC-TCR interactions. Through DeepTR, we hope to serve the community to push forward our understandings of mechanism of T cell-mediated immunity and yield new insight in both personalized immune treatment and development of targeted vaccines.
Deep learning framework of DeepTR models and web server. A. A deep generative model for learning TCR embeddings by LSTM autoencoder. We encoded >20,000 paired α/β TCR sequence data and their six complementary determining regions (CDRs) into high-dimensional physicalchemical feature representations to improved TCR repertoires featurization. B. For embedding of pMHCs, based on over based on more than 200, 000 quantitative binding affinity measurements, we constructed a new pan-allele MHC class I predictor through the deep neural network. The input of this model is the MHC class I sequences and the antigen peptides. The layers before output should contain important information regarding the overall structure of the pMHC complex, which could provide high-quality MHC-antigen co-embedding. C. We leveraged the trained numeric vector encodings of TCRs and pMHCs for learning the pairing between them. We constructed a fully connected deep-learning network based on the output of these two submodels, leading to a final layer with a single neuron for predicting the pairing. D. The web portal implemented by DeepTR models.