What question did this study set out to answer?

The aim is to create a fast and accurate framework for predicting peptide-MHC binding affinities and generating 3D structures.

April 4, 2026Open Access

A high-speed attention network for MHC-bound peptide identification and 3D modeling

Key Points

The aim is to create a fast and accurate framework for predicting peptide-MHC binding affinities and generating 3D structures.
Developed the SwiftMHC framework using deep learning methods trained on synthetic data.
Predicted binding affinities for peptide-MHC using a single A100 GPU in batch mode.
Generated all-atom 3D structures with a focus on specific HLA alleles.
Achieved binding affinity predictions in 0.009 seconds per case.
Maintained a median Cα-RMSD of 1.32 Å compared to crystallographic benchmarks.
Outperformed or matched state-of-the-art methods but at lower computational costs.

Abstract

We developed SwiftMHC, an ultra-fast and accurate structure-based framework for peptide-MHC (pMHC) modeling and binding affinity prediction. Using task-specific deep learning trained on physics-derived synthetic data, SwiftMHC predicts pMHC binding affinities in 0.009 s per case on a single A100 GPU when running in batch mode, offering improved speed compared with leading sequence-based tools such as netMHCpan and MHCflurry while maintaining competitive accuracy. In addition, SwiftMHC generates all-atom 3D pMHC structures with a median Cα-RMSD of 1.32 Å against crystallographic benchmarks, matching or exceeding state-of-the-art methods such as AlphaFold2-finetune but at a lower computational cost. Optimized for HLA-A∗02:01 9-mer peptides but readily extensible to other alleles, SwiftMHC unites structural insight with high-throughput scalability to accelerate safe and effective epitope discovery in cancer immunotherapy.

A high-speed attention network for MHC-bound peptide identification and 3D modeling

Key Points

Abstract

Cite This Study