Email updates

Keep up to date with the latest news and content from BMC Bioinformatics and BioMed Central.

This article is part of the supplement: UT-ORNL-KBRIN Bioinformatics Summit 2010

Open Access Open Badges Poster presentation

Understanding molecular recognition and epitope prediction from Information Theoretic approach

Indranil Mitra1* and Yan Cui2

Author Affiliations

1 Department of Mathematical Sciences, Clemson University, E-006 Martin Hall, Clemson, SC 29631, USA

2 Department of Molecular Sciences & Center of Integrative and Translational Genomics, University of Tennessee Health Science Center, 858 Madison Ave. Memphis, TN 38163, USA

For all author emails, please log on.

BMC Bioinformatics 2010, 11(Suppl 4):P22  doi:10.1186/1471-2105-11-S4-P22

The electronic version of this article is the complete one and can be found online at:

Published:23 July 2010

© 2010 Mitra and Cui; licensee BioMed Central Ltd.


Cellular immunity is dependent on T-cell recognition of peptide/major histocompatibility complex (MHC) and is a critical molecular recognition component [1]. A large class of bioinformatics tools facilitates the identification of T-cell epitopes to specific MHC alleles. However, not all peptide residues contribute equally or are relevant to binding due to polymorphism of genes encoding MHC, making development of statistical methods difficult. Information Theory has proved to be one of the most universal mathematical theories that governs virtually all processes [2]. The success of this approach in analyzing a huge range of engineering, technological and natural processes is impressive. In Molecular Biology the applications have been very successful at the sequence level, many sequence comparison and binding site identification methods now boasts a sound information theoretic foundation.

Materials and methods

In this work we have developed a mathematical formalism for applying information theory in identifying an explicit computational strategy and developing algorithms for the study of peptide/MHC interactions through epitope predictions. A sampling method has been initiated to circumvent the binding problem. Comparisons have been made with existing Machine Learning Methods and a validation of the efficiency of the model may be tested [3,4].The results will have significant impact for understanding the immune system and for rational drug design [5].


This work was partially supported by DOD grant W81XHW-05-01-0227 received by YC. Authors would also like to thank Dr. IrisAntes, Technical University, Munich for helpful discussions.


  1. Lin HH, Zhang GL, Tongchusak S, Reinherzand EL, Brusic V: Evaluation of MHC-II peptide binding prediction servers: applications for vaccine research.

    BMC Bioinformatics 2008, 9(Suppl 12):S12-S22. PubMed Abstract | BioMed Central Full Text | PubMed Central Full Text OpenURL

  2. Adami C: Physics of Life Reviews.

    2004, 1.

  3. O'Brien PJ, Herschlag D: Catalytic promiscuity and the evolution of new enzymatic activities.

    Chem Biol 1999, 6:R91-R105. PubMed Abstract | Publisher Full Text OpenURL

  4. Tong JC, Tan TW, Ranganathan S: Methods and Protocols for Prediction of Immunogenic Epitopes.

    Brief Bioinform 2007, 8(2):96-108. PubMed Abstract | Publisher Full Text OpenURL

  5. Wang P, Sidney J, Dow C, Moth B, Sette A, Peters B: A Systematic Assessment of MHC Class II Peptide Binding Predictions and Evaluation of a Consensus Approach.

    PLoS Comput Biol 2008, 4(4):e1000048. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL