Structural prediction of peptides binding to MHC class I molecules

Area of Science:

• Immunology
• Structural Biology
• Computational Chemistry

Background:

• Peptide binding to MHC class I (MHCI) is crucial for adaptive immunity.
• Accurate prediction of peptide-MHCI interactions is vital for peptide vaccine design.
• Existing sequence-based algorithms lack structural information and accuracy.

Purpose of the Study:

• To present PePSSI, a novel algorithm for predicting peptide structure bound to MHCI molecules.
• To incorporate explicit water molecules at the peptide-MHC interface for improved accuracy.
• To assess the correlation between structural features and peptide binding affinity.

Main Methods:

• PePSSI combines peptide backbone conformation sampling with flexible MHC side-chain movements.
• Explicit water molecules were included at the peptide-MHC interface.
• The algorithm was validated against X-ray crystallography data for eight peptide-HLA-A2 complexes.
• Binding conformations for 266 peptides with known affinities were predicted.

Main Results:

• PePSSI achieved accurate predictions with RMSD values between 1.301 and 2.475 Å compared to X-ray data.
• Higher peptide binding affinity correlated positively with peptide-MHC contacts.
• Increased interfacial water molecules negatively correlated with binding affinity.
• Results align with the hydrophobic nature of the HLA-A2 binding interface.

Conclusions:

• PePSSI enables rapid and accurate prediction of peptide-MHCI binding conformations.
• The algorithm's structural insights can aid in estimating MHCI-peptide binding affinity.
• PePSSI offers a valuable tool for peptide vaccine development and immunoinformatics.