United States
£ GBP
Proteins lie at the heart of almost all biological processes and have an incredibly wide range of activities. Central to the function of all proteins is their ability to adopt, stably or sometimes transiently, structures that allow for interaction with other molecules. An understanding of a protein's structure can therefore lead us to a much improved picture of its molecular function(s). This realisation has been a prime motivation of recent Structural Genomics projects, involving large-scale experimental determination of protein structures, often those of proteins about which little is known of function. These initiatives have, in turn, stimulated the massive development of novel methods for prediction of protein function from structure.
Since model structures may also take advantage of new function prediction algorithms, the first part of the book deals with the various ways in which protein structures may be predicted or inferred, including specific treatment of membrane and intrinsically disordered proteins. A detailed consideration of all current structure-based function prediction methodologies forms the second part of this book, which concludes with two chapters, focusing specifically on case studies, designed to illustrate the real-world application of these methods. With bang up-to-date texts from world experts, and abundant links to publicly available resources, this book will be invaluable to anyone who studies proteins and the endlessly fascinating relationship between their structure and function.
Table of ContentsPart 1: Generating and inferring structures 1 Ab initio protein structure prediction1.1 Introduction1.2 Energy functions1.2.1 Physics-based energy functions1.2.2 Knowledge-based energy function combined with fragments1.3 Conformational search methods1.3.1 Monte Carlo simulations1.3.2 Molecular dynamics1.3.3 Genetic algorithm1.3.4 Mathematical optimization1.4 Model selection1.4.1 Physics-based energy function1.4.2 Knowledge-based energy function1.4.3 Sequence-structure compatibility function1.4.4 Clustering of decoy structure1.5 Remarks and discussion2 Fold Recognition 2.1 Introduction2.1.1 The importance of blind trials: the CASP competition2.1.2 Ab initio structure prediction versus homology modelling2.1.3 The limits of fold space2.1.4 A note on terminology: 'threading' and 'fold recognition'2.2 Threading2.2.1 Knowledge-based potentials2.2.2 Finding an alignment2.2.3 Heuristics for alignment2.3 Remote homology detection without threading2.3.1 Using predicted structural features2.3.2 Sequence profiles and hidden Markov models2.3.3 Fold Classification and Support Vector Machines2.3.4 Consensus approaches2.3.5 Traversing the homology network2.4 Alignment accuracy, model quality and statistical significance2.4.1 Algorithms for alignment generation and assessment2.4.2 Estimation of statistical significance2.5 Tools for fold recognition on the web2.6 The future3 Comparative protein structure modelling3.1 Introduction3.1.1 Structure determines function3.1.2 Sequences, structures, structural genomics3.1.3 Approaches to protein structure prediction3.2 Steps in comparative protein structure modelling3.2.1 Searching for structures related to the target sequence3.2.2 Selecting templates3.2.3 Sequence to structure alignment3.2.4 Model building3.2.5 Model evaluation3.3 Performance of comparative modelling3.3.1 Accuracy of methods3.3.2 Errors in comparative models3.4 Applications of comparative modelling3.4.1 Modelling of individual proteins3.4.2 Comparative modelling and the Protein Structure Initiative3.5 Summary4 Membrane protein structure prediction4.1 Introduction4.2 Structural classes4.2.1 Alpha-helical bundles4.2.2 Beta-barrels4.3 Membrane proteins are difficult to crystallise4.4 Databases4.5 Multiple sequence alignments4.6 Transmembrane protein topology prediction4.6.1 Alpha-helical proteins4.6.2 Beta-barrel proteins4.6.3 Whole genome analysis4.6.4 Data sets, homology, accuracy and cross-validation4.7 3D structure prediction4.8 Future developments5 Bioinformatics approaches to the structure and function of intrinsically disordered proteins5.1 The concept of protein disorder5.2 Sequence features of IDPs5.2.1 The unusual amino acid composition of IDPs5.2.2 Sequence patterns of IDPs5.2.3 Low sequence complexity and disorder5.3 Prediction of disorder5.3.1 Prediction of low-complexity regions5.3.2 Charge-hydropathy plot5.3.3 Propensity-based predictors5.3.4 Predictors based on the lack of secondary structure5.3.5 Machine learning algorithms5.3.6 Prediction based on contact potentials5.3.7 A reduced alphabet suffices to predict disorder5.3.8 Comparison of disorder prediction methods5.4 Functional classification of IDPs5.4.1 Gene Ontology-based functional classification of IDPs5.4.2 Classification of IDPs based on their mechanism of action5.4.3 Function-related structural elements in IDPs5.5 Prediction of the function of IDPs5.5.1 Correlation of disorder pattern and function5.5.2 Predicting short recognition motifs in IDRs5.5.3 Prediction of MoRFs5.5.4 Combination of information on sequence and disorder: phosphorylation sites and CaM binding motifs5.5.5 Flavours of disorder5.6 Limitations to IDP function prediction5.6.1 Rapid evolution of IDPs5.6.2 Sequence independence of function and fuzziness5.6.3 Good news: conservation and disorder5.7 ConclusionsPart 2: From structures to functions6 Function diversity within folds and superfamilies6.1 Defining function6.2 From fold to function6.2.1 Definition of a fold6.2.2 Prediction of function using fold relationships6.3 Function diversity between homologous proteins6.3.1 Definitions6.3.2 Evolution of protein superfamilies6.3.3 Function divergence during protein evolution6.4 Conclusion7 Predicting protein function from surface properties7.1 Surface descriptors7.1.1 The van der Waals surface7.1.2 Molecular surface (solvent excluded surface) 7.1.3 The solvent accessible surface7.2 Surface properties7.2.1 Hydrophobicity7.2.2 Electrostatics properties7.2.3 Surface conservation7.3 Function predictions using surface properties7.3.1 Hydrophobic surface7.3.2 Electrostatic surface7.3.3 Surface conservation7.3.4 Combining surface properties for function prediction7.4 Protein-ligand interactions7.4.1 Properties of protein-ligand interactions7.4.2 Predicting binding site locations7.4.3 Predictions of druggability7.4.4 Annotation of ligand binding sites7.5 Protein-protein interfaces7.5.1 Properties of protein-protein interfaces7.5.2 Hot-spot regions in protein interfaces7.5.3 Predictions of interface location7.6 Summary8 3D Motifs8.1 Background and significance8.1.1 What is function? 8.1.2 3D Motifs: Definition and scope8.2 Overview of Methods8.2.1 Motif discovery8.2.2 Motif description and matching8.2.3 Interpretation of results8.3 Specific Methods8.3.1 User-defined motifs8.3.2 Motif discovery8.4 Related methods8.4.1 Hybrid (point-surface) descriptions8.4.2 Single-point-centred descriptions8.5 Docking for functional annotation8.6 Discussion8.7 Conclusions9 Protein dynamics: from structure to function9.1 Molecular dynamics simulations9.1.1 Principles and approximations9.1.2 Applications9.1.3 Limitations9.2 Principal component analysis9.3 Collective coordinate sampling algorithms9.3.1 Essential dynamics9.3.2 TEE-REX9.4 Methods for functional mode prediction9.4.1 Normal mode analysis9.4.2 Elastic network models9.4.3 CONCOORD9.5 Summary and outlook10 Integrated servers for structure-informed function prediction10.1 Introduction10.1.1 The problem of predicting function from structure10.1.2 Structure-function prediction methods10.2 ProKnow10.2.1 Fold matching10.2.2 3D motifs10.2.3 Sequence homology10.2.4 Sequence motifs10.2.5 Protein interactions10.2.6 Combining the predictions10.2.7 Prediction success10.3 ProFunc10.3.1 ProFunc's structure-based methods10.3.2 Assessment of the structural methods10.4 Conclusion11. Case studies: Function predictions of structural genomics results11.1 Introduction11.2 Large scale function prediction case studies11.3 Some specific examples11.4 Community annotation11.5 Conclusions12. Prediction of protein function from theoretical models12.1 Background12.2 Protein models as a community resource12.2.1 Model quality12.2.2 Databases of models12.3 Accuracy and added value of model-derived properties12.3.1 Implementation12.4 Practical application12.4.1 Plasticity of catalytic site residues12.4.2 Mutation mapping12.4.3 Protein complexes12.4.4 Function predictions from template-free models12.4.5 Prediction of ligand specificity12.4.6 Structure modelling of alternatively spliced isoforms12.4.7 From broad function to molecular details12.5 What next? Index
