Immunological Bioinformatics Sorin Istrail, Pavel Pevzner, and Michael Waterman, editors Computational molecular biology is a new discipline, bringing together computational, statistical, experimental, and technological methods, which is energizing and dramatically accelerating the discovery of new technologies and tools for molecular biology. The MIT Press Series on Computational Molecular Biology is intended to provide a unique and effective venue for the rapid publication of monographs, textbooks, edited collections, reference works, and lecture notes of the highest quality. Computational Molecular Biology: An Algorithmic Approach Pavel A. Pevzner, 2000 Computational Methods for Modeling Biochemical Networks James M. Bower and Hamid Bolouri, editors, 2001 Current Topics in Computational Molecular Biology Tao Jiang, Ying Xu, and Michael Q. Zhang, editors, 2002 Gene Regulation and Metabolism: Postgenomic Computation Approaches Julio Collado-Vides, editor, 2002 Microarrays for an Integrative Genomics Isaac S. Kohane, Alvin Kho, and Atul J. Butte, 2002 Kernel Methods in Computational Biology Bernhard Schölkopf, Koji Tsuda and Jean-Philippe Vert, editors, 2004 An Introduction to Bioinformatics Algorithms Neil C. Jones and Pavel A. Pevzner, 2004 Immunological Bioinformatics Ole Lund, Morten Nielsen, Claus Lundegaard, Can Ke¸smirand Søren Brunak, 2005 Ontologies for Bioinformatics Kenneth Baclawski and Tianhua Niu, 2005 Immunological Bioinformatics Ole Lund Morten Nielsen Claus Lundegaard Can Ke¸smir Søren Brunak The MIT Press Cambridge, Massachusetts London, England c 2005 Massachusetts Institute of Technology All rights reserved. No part of this book may be reproduced in any form by any electronic or mechanical means (including photocopying, recording, or information storage and retrieval) without permission in writing from the publisher. MIT press books may be purchased at special quantity discounts for business or sales promotional use. For information please email spe- [email protected] or write to Special Sales Department, The MIT press, 55 Hayward Street, Cambridge, MA 02142. This book was set in Lucida by the authors and was printed and bound in the United States of America. Library of Congress Cataloging-in-Publication Data Immunological bioinformatics / Ole Lund .. [et al.]. p. cm. — (Computational molecular biology) Includes bibliographical references and index. ISBN 0-262-12280-4 (alk. paper) 1. Immunoinformatics. I. Lund, Ole. II. Series. QR182.2.I46I465 2005 571.9’6’0285-dc22 2005042806 Contents Preface ix 1 Immune Systems and Systems Biology 1 1.1 Innate and Adaptive Immunity in Vertebrates 10 1.2 Antigen Processing and Presentation 11 1.3 Individualized Immune Reactivity 14 2 Contemporary Challenges to the Immune System 17 2.1 Infectious Diseases in the New Millennium 17 2.2 Major Killers in the World 17 2.3 Childhood Diseases 20 2.4 Clustering of Infectious Disease Organisms 22 2.5 Biodefense Targets 28 2.6 Cancer 30 2.7 Allergy 30 2.8 Autoimmune Diseases 31 3 Sequence Analysis in Immunology 33 3.1 Sequence Analysis 33 3.2 Alignments 34 3.3 Multiple Alignments 50 3.4 DNA Alignments 52 3.5 Molecular Evolution and Phylogeny 53 3.6 Viral Evolution and Escape: Sequence Variation 55 3.7 Prediction of Functional Features of Biological Sequences 59 4 Methods Applied in Immunological Bioinformatics 67 4.1 Simple Motifs, Motifs and Matrices 67 4.2 Information Carried by Immunogenic Sequences 70 4.3 Sequence Weighting Methods 73 4.4 Pseudocount Correction Methods 75 v vi Contents 4.5 Weight on Pseudocount Correction 77 4.6 Position Specific Weighting 77 4.7 Gibbs Sampling 78 4.8 Hidden Markov Models 82 4.9 Artificial Neural Networks 89 4.10 Performance Measures for Prediction Methods 97 4.11 Clustering and Generation of Representative Sets 100 5 DNA Microarrays in Immunology 101 5.1 DNA Microarray Analysis 101 5.2 Clustering 104 5.3 Immunological Applications 106 6 Prediction of Cytotoxic T Cell (MHC Class I) Epitopes 109 6.1 Background and Historical Overview of Methods for Pep- tide MHC Binding Prediction 110 6.2 MHC Class I Epitope Binding Prediction Trained on Small Data Sets 112 6.3 Prediction of CTL Epitopes by Neural Network Methods 118 6.4 Summary of the Prediction Approach 131 7 Antigen Processing in the MHC Class I Pathway 133 7.1 The Proteasome 133 7.2 Evolution of the Immunosubunits 135 7.3 Specificity of the (Immuno)Proteasome 137 7.4 Predicting Proteasome Specificity 141 7.5 Comparison of Proteasomal Prediction Performance 145 7.6 Escape from Proteasomal Cleavage 147 7.7 Post-Proteasomal Processing of Epitopes 148 7.8 Predicting the Specificity of TAP 151 7.9 Proteasome and TAP Evolution 152 8 Prediction of Helper T Cell (MHC Class II) Epitopes 155 8.1 Prediction Methods 156 8.2 The Gibbs Sampler Method 157 8.3 Further Improvements of the Approach 170 9 Processing of MHC Class II Epitopes 173 9.1 Enzymes Involved in Generating MHC Class II Ligands 174 9.2 Selective Loading of Peptides to MHC Class II Molecules 177 9.3 Phylogenetic Analysis of the Lysosomal Proteases 178 9.4 Signs of the Specificities of Lysosomal Proteases on MHC Class II Epitopes 180 Contents vii 9.5 Predicting the Specificity of Lysosomal Enzymes 180 10 B Cell Epitopes 185 10.1 Affinity Maturation 186 10.2 Recognition of Antigen by B cells 189 10.3 Neutralizing Antibodies 199 11 Vaccine Design 201 11.1 Categories of Vaccines 202 11.2 Polytope Vaccine: Optimizing Plasmid Design 205 11.3 Therapeutic Vaccines 207 11.4 Vaccine Market 211 12 Web-Based Tools for Vaccine Design 213 12.1 Databases of MHC Ligands 213 12.2 Prediction Servers 215 13 MHC Polymorphism 221 13.1 What Causes MHC Polymorphism? 221 13.2 MHC Supertypes 223 14 Predicting Immunogenicity: An Integrative Approach 241 14.1 Combination of MHC and Proteasome Predictions 242 14.2 Independent Contributions from TAP and Proteasome Predictions 243 14.3 Combinations of MHC, TAP, and Proteasome Predictions 245 14.4 Validation on HIV Data Set 249 14.5 Perspectives on Data Integration 250 References 252 viii Contents Preface The immune responses are extraordinarily complex, involving the dynamic in- teraction of a wide array of tissues, cells, and molecules. Immunology has traditionally been a qualitative science describing the cellular and molecular components of the immune system and their functions. The traditional ap- proaches are by and large reductionist, avoiding complexity, but providing detailed knowledge of a single event, cell, or molecular entity. The sequencing of the human genome, in concert with emerging genomic and proteomic tech- nologies, changed the way of studying the immune system drastically. The immunologists are now, maybe for the first time, aiming to provide a compre- hensive description of the complex immunological processes. Generation of huge amounts of data made it clear that this goal cannot be achieved without using powerful computational approaches. Wherever cellular life occurs, viruses are also found. The immune sys- tems are evolved to defend the organism against these intruders. Since viruses evade or interfere with specific cellular pathways to escape immune responses, knowledge of viral genome sequences has helped, in some cases, fundamental understanding of host biology. Studying host-virus interactions at the level of single gene effects, however, fails to produce a global systems level under- standing. This should now be achievable in the context of complete host and pathogen genome sequences. So again, understanding host-pathogen interac- tions calls for a close collaboration between microbiology and immunology at the systems-level. Immunological bioinformatics is the research field that applies informatics techniques to generate a systems-level view of the immune system. The long- term goal of the research is to establish an in silico immune system. This may be done in a stepwise fashion where models are developed for the different components of the immune system. These models can be combined and may help to understand diseases, and develop therapies, vaccines, and diagnostic tools for treatment of major killers such as AIDS, malaria, and cancer. The immune system does not react to entire pathogens but rather to short fragments (epitopes) of proteins from pathogens. A major branch of immuno- x Preface logical bioinformatics is dedicated to identifying these immunogenic regions in a broad sense. This book reviews the current state of the art of this branch and other (related) immunological bioinformatics research. Audience and Prerequisites The book is aimed at both students and more advanced researchers with di- verse backgrounds. We have tried to provide a succinct description of the main biological concepts and problems for readers with a strong background in mathematics, statistics, and computer science. Likewise, the book is tailored to biologists and biochemists who will often know more about the biological problems than the text explains, but need some help in understanding the new data-driven algorithms in the context of biological data. It should in principle provide enough insights while remaining sufficiently simple for the reader to be able to implement the algorithms described, or adapt them to a particular problem. Content and General Outline of the Book We have tried to write a book that is more or less self-contained. The bioin- formatics methods are first explained in an intuitive way, and later we go into more detail of the mathematics lying behind them. Only chapter 4 is ded- icated to a detailed description of the basic methods. A significant portion of the book is built on material taken from articles we have written over the years, as well as from tutorials given at several conferences, including the ISMB (Intelligent Systems for Molecular Biology) conferences, courses given at the Technical University of Denmark and Utrecht University. In each chapter we have tried to show the interesting biological insights gained from the bioinformatics approach. This, we hope demonstrates how and why bioinformatics can be used to understand the complexity of the im- mune system.