Computational design of HLA class I superbinders for broad T cell immunogenicity
Peer, E.; Cohen-Lavi, L.; Sette, A.; Sidney, J.; Hertz, T.
Show abstract
Human leukocyte antigen (HLA) class I molecules are highly polymorphic, restricting peptide binding to narrow sequence subsets. Designing peptides that bind multiple HLA supertypes-- termed superbinders--offers a promising strategy for broad-spectrum T-cell vaccines and immunotherapies. Here we present superHLA, a computational framework that combines Markov Chain Monte Carlo optimization with state-of-the-art MHC binding predictors to design synthetic 9-mer peptides with broad HLA-binding profiles. Using superHLA, we generated over 190,000 candidate superbinders predicted to bind 8-12 HLA class I alleles across distinct supertypes. A multi-tier filtering pipeline--incorporating sequence clustering, synthesis feasibility, cross-predictor validation, and self-peptidome exclusion--yielded a final panel of 100 peptides for experimental testing. Of these, 21 bound [≥]4 supertypes in vitro, including one that bound 9. Superbinders displayed distinct anchor residue preferences and showed minimal similarity to human peptides. These results suggest that HLA superbinders are more abundant than previously recognized and can be rationally designed at scale. This approach supports development of pan-HLA immunogens with broad population coverage and potential applications in vaccine design, neoantigen discovery, and immunotherapy.
Matching journals
The top 7 journals account for 50% of the predicted probability mass.
Similar papers in this journal
- The immunogenic potential of recurrent cancer drug resistance mutations: an in silico study 95%
- Influence of HLA class II polymorphism on predicted cellular immunity against SARS-CoV-2 at the population and individual level 95%
- Feature Selection Enhances Peptide Binding Predictions for TCR-Specific Interactions 95%
Similar papers in this journal
- THLANet: A Deep Learning Framework for Predicting TCR-pHLA Binding in Immunotherapy Applications 94%
- Exploring the Potential of Structure-Based Deep Learning Approaches for T cell Receptor Design 94%
- An Integrated Approach to the Characterization of Immune Repertoires Using AIMS: An Automated Immune Molecule Separator 94%
Similar papers in this journal
Similar papers in this journal
- NetTCR 2.2 - Improved TCR specificity predictions by combining pan- and peptide-specific training strategies, loss-scaling and integration of sequence similarity 94%
- Structure-based prediction of T cell receptor:peptide-MHC interactions 94%
- Biochemical Patterns of Antibody Polyreactivity Revealed Through a Bioinformatics-Based Analysis of CDR Loops 94%
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.