Related Experiment Video
Updated: May 10, 2026
09:32
Immunopeptidomics: Isolation of Mouse and Human MHC Class I- and II-Associated Peptides for Mass Spectrometry Analysis
Published on: October 15, 2021
Protein language models accurately predict polymorphic peptide-modulated NK cell receptor-HLA class I interaction
Abdallah AlShafey1, Madeline Nelson2,3, Mubasher Hassan1
1Steve and Cindy Rasmussen Institute for Genomic Medicine, Abigail Wexner Research Institute, Nationwide Children's Hospital, Columbus, OH 43205, USA.
Science Advances
|May 8, 2026
Summary
Predicting peptides that bind to killer-cell immunoglobulin-like receptors (KIRs) is crucial for personalized immunotherapy. A new AI model accurately identifies these KIR-binding peptides, advancing cancer and infectious disease treatments.
Area of Science:
- Immunology
- Computational Biology
- Structural Biology
Background:
- Killer-cell immunoglobulin-like receptors (KIRs) regulate natural killer cell function.
- KIRs interact with human leukocyte antigen class I (HLA-I) molecules, influenced by bound peptides.
- These interactions are vital in immune responses to infections, inflammation, and cancer.
Purpose of the Study:
- To develop a computational tool for identifying peptides that bind to KIRs.
- To leverage foundation protein language models for predicting KIR-HLA-peptide interactions.
- To enable personalized immunotherapy strategies by identifying disease-specific peptides.
Main Methods:
- Training foundation protein language models on existing KIR-binding peptide-HLA complex datasets.
- Utilizing sequence data from interacting molecules to predict peptide binding.
- Evaluating model performance using area under the receiver operator characteristic (AUROC) curves.
Main Results:
- The developed tool achieved high prediction accuracy (AUROC >0.8) for most inhibitory KIRs.
- The model demonstrated good performance (AUROC >0.7) for peptides associated with HIV and HCV infections.
- Successful prediction of KIR-binding peptides from viral and cancer sources.
Conclusions:
- The AI model offers a valuable tool for identifying disease-associated peptides.
- This work advances understanding of immune regulation and biophysical binding factors.
- The findings pave the way for developing KIR-specific immunotherapies against cancer and infections.
