DNA Methylation Microarrays: Experimental Design and Statistical Analysis
Purchasing Options
Features
Summary
Providing an interface between dry-bench bioinformaticians and wet-lab biologists, DNA Methylation Microarrays: Experimental Design and Statistical Analysis presents the statistical methods and tools to analyze high-throughput epigenomic data, in particular, DNA methylation microarray data. Since these microarrays share the same underlying principles as gene expression microarrays, many of the analyses in the text also apply to microarray-based gene expression and histone modification (ChIP-on-chip) studies.
After introducing basic statistics, the book describes wet-bench technologies that produce the data for analysis and explains how to preprocess the data to remove systematic artifacts resulting from measurement imperfections. It then explores differential methylation and genomic tiling arrays. Focusing on exploratory data analysis, the next several chapters show how cluster and network analyses can link the functions and roles of unannotated DNA elements with known ones. The book concludes by surveying the open source software (R and Bioconductor), public databases, and other online resources available for microarray research.
Requiring only limited knowledge of statistics and programming, this book helps readers gain a solid understanding of the methodological foundations of DNA microarray analysis.
Table of Contents
Preface
Applied Statistics
Descriptive statistics
Inferential statistics
DNA Methylation Microarrays and Quality Control
DNA methylation microarrays
Workflow of methylome experiment
Image analysis
Visualization of raw data
Reproducibility
Experimental Design
Goals of experiment
Reference design
Balanced block design
Loop design
Factorial design
Time course experimental design
How many samples/arrays are needed?
Appendix
Data Normalization
Measure of methylation
The need for normalization
Strategy for normalization
Two-color CpG island microarray normalization
Oligonucleotide arrays normalization
Normalization using control sequences
Appendix
Significant Differential Methylation
Fold change
Linear model for log-ratios or log-intensities
t test for contrasts
F test for joint contrasts
P-value adjustment for multiple testing
Modified t and F tests
Significant variation within and between groups
Significant correlation with a covariate
Permutation test for bisulfite sequence data
Missing data values
Appendix
High-Density Genomic Tiling Arrays
Normalization
Wilcoxon test in a sliding window
Boundaries of methylation regions
Multiscale analysis by wavelets
Unsupervised segmentation by hidden Markov model
Principal component analysis and biplot
Cluster Analysis
Measure of dissimilarity
Dimensionality reduction
Hierarchical clustering
K-means clustering
Model-based clustering
Quality of clustering
Statistical significance of clusters
Reproducibility of clusters
Repeated measurements
Statistical Classification
Feature selection
Discriminant function
K-nearest neighbor
Performance assessment
Interdependency Network of DNA Methylation
Graphs and networks
Partial correlation
Dependence networks from DNA methylation microarrays
Network analysis
Time Series Experiment
Regulatory networks from microarray data
Dynamic model of regulation
A penalized likelihood score for parsimonious model
Optimization by genetic algorithms
Online Annotations
Gene centric resources
PubMeth: A cancer methylation database
Gene Ontology
Kyoto Encyclopedia of Genes and Genomes
UniProt/Swiss-Prot protein knowledgebase
The International HapMap Project
UCSC human genome browser
Public Microarray Data Repositories
Epigenetics Society
Microarray Gene Expression Data Society
Minimum Information About a Microarray Experiment
Public repositories for high-throughput arrays
Open Source Software for Microarray Data Analysis
R: A language and environment for statistical computing and graphics
Bioconductor
References
Index
Editorial Reviews
I found the book to be very informative and a timely introduction to the issues related to designing and analyzing array-based methylation experiments. … it provides a solid grounding and serves as a good reference book for any statistician venturing into this field.
—Sarah Bujac, Pharmaceutical Statistics, 2011, 10
…a useful presentation of four detailed, well-written parts concerning techniques in the analysis of high throughput epigenomic data … a consistent and self-contained overview on important fundamental and modern procedures used by researchers in biology, bioinformatics, experimental designs …The book is of great interest to research workers who use the above-mentioned procedures in experimental design and deep analysis of epigenomic data with sound statistics.
—Cryssoula Ganatsiou, Zentralblatt MATH 1172
…This book is a helpful guide for researchers and students with an interest in performing genomic studies using high-throughput microarrays. … A wide range of useful data analysis tools are covered … Other strengths throughout the book include the discussion of experimental design, the mention of software for certain analyses, and the inclusion of more advanced methods such as wavelets and genetic algorithms. … Overall, this book gives a nice summary of methods used for the analysis of hybridization-based microarray data. …
—Biometrics, March 2009
