Yona Introduction to Computational Proteomics

1. Auflage 2013
ISBN: 978-1-4200-1077-0
Verlag: Taylor & Francis
Format: PDF
Kopierschutz: Adobe DRM (»Systemvoraussetzungen)

E-Book, Englisch, 767 Seiten

Reihe: Chapman & Hall/CRC Mathematical & Computational Biology

Introduction to Computational Proteomics introduces the field of computational biology through a focused approach that tackles the different steps and problems involved with protein analysis, classification, and meta-organization. The book starts with the analysis of individual entities and works its way through the analysis of more complex entities, from protein families to interactions, cellular pathways, and gene networks.

The first part of the book presents methods for identifying the building blocks of the protein space, such as motifs and domains. It also describes algorithms for assessing similarity between proteins based on sequence and structure analysis as well as mathematical models, such as hidden Markov models and support vector machines, that are used to represent protein families and classify new instances.

The second part covers methods that investigate higher order structure in the protein space through the application of unsupervised learning algorithms, such as clustering and embedding. The book also explores the broader context of proteins. It discusses methods for analyzing gene expression data, predicting protein-protein interactions, elucidating cellular pathways, and reconstructing gene networks.

This book provides a coherent and thorough introduction to proteome analysis. It offers rigorous, formal descriptions, along with detailed algorithmic solutions and models. Each chapter includes problem sets from courses taught by the author at Cornell University and the Technion. Software downloads, data sets, and other material are available at biozon.org

Autoren/Hrsg.

Yona, Golan

Fachgebiete

• Mathematik | Informatik Mathematik Numerik und Wissenschaftliches Rechnen Angewandte Mathematik, Mathematische Modelle
• Mathematik | Informatik Mathematik Stochastik
• Technische Wissenschaften Verfahrenstechnik | Chemieingenieurwesen | Biotechnologie Biotechnologie

Inhaltsverzeichnis

PART I: THE BASICS

What Is Computational Proteomics?

The complexity of living organisms

Proteomics in the modern era

The main challenges in computational proteomics

Basic Notions in Molecular Biology

The cell structure of organisms

It all starts from the DNA

Proteins

From DNA to proteins

Protein folding—from sequence to structure

Evolution and relational classes in the protein space

Sequence Comparison

Alignment of sequences
Heuristic algorithms for sequence comparison

Probability and statistics of sequence alignments
Scoring matrices and gap penalties
Distance and pseudo-distance functions for proteins

Further reading

Conclusions
Appendix: performance evaluation
Appendix: basic concepts in probability

Multiple Sequence Alignment, Profiles, and Partial Order Graphs

Dynamic programming in N dimensions

Classical heuristic methods
MSA representation and scoring
Profile analysis
Iterative and progressive alignment
Transitive alignment

Partial order alignment
Further reading

Conclusions

Motif Discovery

Introduction

Model-based algorithms
Searching for good models: Gibbs sampling and MEME
Combinatorial approaches
Further reading

Conclusions

Appendix: the expectation-maximization algorithm

Markov Models of Protein Families

Introduction

Markov models
Main applications of hidden Markov models (the evaluation and decoding problems)
Learning HMMs from data
Higher order models, codes and compression
Variable order Markov models
Further reading

Conclusions

Classifiers and Kernels

Generative models vs discriminative models

Classifiers and discriminant functions
Applying SVMs to protein classification
Decision trees
Further reading

Conclusions
Appendix

Protein Structure Analysis
Introduction

Structure prediction—the protein folding problem
Structure comparison
Generalized sequence profiles—integrating secondary structure with sequence information

Further reading

Conclusions

Appendix

Protein Domains
Introduction

Domain detection
Learning domain boundaries from multiple features
Testing domain predictions
Multi-domain architectures
Further reading

Conclusions

Appendix

PART II: PUTTING ALL THE PIECES TOGETHER
Clustering and Classification

Introduction

Clustering methods

Vector-space clustering algorithms
Graph-based clustering algorithms
Collaborative clustering
Spectral clustering algorithms
Markovian clustering algorithms
Cluster validation and assessment
Clustering proteins
Further reading

Conclusions

Appendix

Embedding Algorithms and Vectorial Representations

Introduction

Structure preserving embedding
Maximal variance embeddings (PCA, SVD)
Distance preserving embeddings (MDS, random projections)
Manifold learning—topological embeddings (IsoMap, LLE, distributional scaling)
Setting the dimension of the host space

Vectorial representations
Further reading

Conclusions

Analysis of Gene Expression Data
Introduction

Microarrays

Analysis of individual genes

Pairwise analysis
Cluster analysis and class discovery
Enrichment analysis
Protein arrays

Further reading

Conclusions

Protein-Protein Interactions

Introduction

Experimental detection of protein interactions
Prediction of protein-protein interactions
Structure-based prediction, protein docking
Sequence-based inference (gene preservation, co-evolution, sequence signatures, and domain-based prediction)
Topological properties of interaction networks
Network motifs
Further reading

Conclusions

Appendices

Cellular Pathways
Introduction
Metabolic pathways
Pathway prediction
Pathway prediction from blueprints
Expression data and pathway analysis
Regulatory networks and modules
Pathway networks and the minimal cell
Further reading
Conclusions
Bayesian Belief Networks

Introduction

Computing the likelihood of observations

Probabilistic inference

Learning the parameters of a Bayesian network

Learning the structure of a Bayesian network
Further reading

Conclusions
References
Problems appear at the end of each chapter.