Quantitative prediction of class I MHC/epitope binding affinity using QSAR modeling derived from amino acid...

Area of Science:

• Immunology
• Computational Biology
• Biochemistry

Background:

• T cell activation relies on peptide-MHC binding, crucial for immune responses.
• Accurate prediction of MHC-epitope binding affinity is vital for efficient epitope screening.
• Current methods can be costly and labor-intensive, necessitating improved predictive tools.

Purpose of the Study:

• To develop a comprehensive quantitative prediction method for MHC-epitope binding affinity.
• To establish robust quantitative structure-activity relationship (QSAR) models for MHC Class I subtypes.
• To utilize amino acid physicochemical properties for predicting binding affinity.

Main Methods:

• Epitopes were characterized using amino acid physicochemical parameters.
• Stepwise regression (STR) was employed for structural variable optimization.
• Multiple linear regression (MLR) was used to build quantitative models for 31 MHC Class I subtypes.

Main Results:

• Developed robust QSAR models for predicting MHC-epitope binding affinity.
• Normalized regression coefficients (NRCs) revealed interaction mechanisms between MHC, epitope, and TCR.
• Identified the contribution of specific amino acids at epitope positions to binding affinity.

Conclusions:

• The STR-MLR models provide a quantitative and mechanistic approach to predict binding affinity.
• These models can guide virtual screening and design of T cell epitopes, particularly CTL epitopes.
• The method offers advantages in terms of clear physicochemical indication, ease of calculation, and performance.