TCRpMHCmodels: Structural modelling of TCR-pMHC class I complexes

Abstract

The interaction between the class I major histocompatibility complex (MHC), the peptide presented by the MHC and the T-cell receptor (TCR) is a key determinant of the cellular immune response. Here, we present TCRpMHCmodels, a method for accurate structural modelling of the TCR-peptide-MHC (TCR-pMHC) complex. This TCR-pMHC modelling pipeline takes as input the amino acid sequence and generates models of the TCR-pMHC complex, with a median Cα RMSD of 2.31 Å. TCRpMHCmodels significantly outperforms TCRFlexDock, a specialised method for docking pMHC and TCR structures. TCRpMHCmodels is simple to use and the modelling pipeline takes, on average, only two minutes. Thanks to its ease of use and high modelling accuracy, we expect TCRpMHCmodels to provide insights into the underlying mechanisms of TCR and pMHC interactions and aid in the development of advanced T-cell-based immunotherapies and rational design of vaccines. The TCRpMHCmodels tool is available at http://www.cbs.dtu.dk/services/TCRpMHCmodels/ .

Figure 1

Flowchart of the computational framework for modelling TCR-pMHC complexes, from the input sequence to the final TCR-pMHC model. The MHC molecule is depicted in blue and the peptide in orange, while the two chains of TCR, α and β, are represented in light and dark grey, respectively.

Figure 2

RMSD accuracy for the different template selection methods. (A) The RMSD for the pMHC complex. (B) The RMSD for the peptide. For each target in the template database, we generate four models using the four different sequence identity thresholds. The OneUnweighted method uses only a single pMHC template with no weights on the sequence identity. The MultiUnweighted also have no weights on the sequence identity but this method uses multiple templates. The OneWeighted method uses only a single pMHC template and a weighted sequence identity. The MultiWeighted method uses the weighted sequence identity and multiple templates. The four different template selection methods are compared with a random baseline (see the Method section for more details). The p-values were obtained using the Wilcoxon signed-rank test.

Figure 3

Benchmark results for the pMHC models generated using the MultiWeighted method. Chothia-Lesk plot showing the RMSD accuracy for the pMHC (orange) and the peptide (red) against the sequence identity to the best template.

Figure 4

The TCR-pMHC RMSD accuracy for the different template selection methods. For each target in the template database we generated four models using the four different sequence identity thresholds and evaluated the generated models using the RMSD for the TCR-pMHC complex. The OneUnweighted method uses only a single TCR-pMHC template with no weights on the sequence identity. The OneWeighted method uses only a single TCR-pMHC template and a weighted sequence identity. The MultiWeighted method uses the weighted sequence identity and multiple templates. The three different template selection methods are compared with a random baseline shown in grey. The p-values were obtained using the Wilcoxon signed-rank test. The RMSD accuracy of the TCR, the pMHC and the peptide are shown in Supplementary Fig. S5.

Figure 5

RMSD accuracy for the TCR-pMHC models generated using the MultiWeighted method. Chothia-Lesk plot showing the RMSD accuracy for the TCR-pMHC (grey), the TCR (blue), the pMHC (orange) and the peptide (red) against the sequence identity to the best template.

Figure 6

Benchmark analysis of the TCR-pMHC models. (A) Shows the TCR-pMHC RMSD accuracy between the models produced by TCRpMHCmodels and TCRFlexDock. (B) Shows the DockQ scores between the models produced by TCRpMHCmodels and TCRFlexDock. The statistical comparison was performed using the Wilcoxon signed-rank test.

Figure 7

CDR accuracy of the TCR-pMHC models from the benchmark analysis. Shows the CDR RMSD accuracy between the TCR-pMHC models produced by the TCRpMHCmodels pipeline and TCRFlexDock protocol, compared to the initial TCR model produced by the TCR-pMHC pipeline.

Figure 8

Database visualization. (A) Number of structures in each template database. (B) The peptide length distribution of structures in the pMHC database. (C) The peptide length distribution of structures in the TCR-pMHC database.