Additive method for the prediction of protein-peptide binding affinity. Application to the MHC class I molecule HLA-A*0201

Irini A. Doytchinova, Martin J. Blythe, Darren R. Flower

Research output: Contribution to journalArticlepeer-review

Abstract

A method has been developed for prediction of binding affinities between proteins and peptides. We exemplify the method through its application to binding predictions of peptides with affinity to major histocompatibility complex class I molecule HLA-A*0201. The method is named “additive” because it is based on the assumption that the binding affinity of a peptide could be presented as a sum of the contributions of the amino acids at each position and the interactions between them. The amino acid contributions and the contributions of the interactions between adjacent side chains and every second side chain were derived using a partial least squares (PLS) statistical methodology using a training set of 420 experimental IC50 values. The predictive power of the method was assessed using rigorous cross-validation and using an independent test set of 89 peptides. The mean value of the residuals between the experimental and predicted pIC50 values was 0.508 for this test set. The additive method was implemented in a program for rapid T-cell epitope search. It is universal and can be applied to any peptide−protein interaction where binding data is known.
Original languageEnglish
Pages (from-to)263-272
Number of pages10
JournalJournal of Proteome Research
Volume1
Issue number3
DOIs
Publication statusPublished - May 2002

Keywords

  • qsar
  • pls
  • free-wilson
  • mhc
  • hla-a*0201

Fingerprint

Dive into the research topics of 'Additive method for the prediction of protein-peptide binding affinity. Application to the MHC class I molecule HLA-A*0201'. Together they form a unique fingerprint.

Cite this