ABSTRACT
Introduction 44
Normalization Methods 44
Per Array Normalization 45
Normalize to a Median or Percentile 45
Normalize to Positive Control Genes 46
Normalize to a Constant Value 46
Per Gene Normalizations 46
Normalization to Specific Samples 46
Normalize to Median 47
Data Analysis 47
Fold Changes in Expression Levels 48
Class Discovery-Unsupervised Learning Methods 50
k-Means Clustering 50
Self-Organizing Maps 51
Principal Component Analysis 51
Hierarchical Trees (Clustering) 54
Class Prediction-Supervised Learning Methods 60
Class Prediction 60
Cross-Validation 61
Validation of Microarray Analysis 61
In silico Validation 61
Laboratory Based Validation 62
Microarray Analysis Software 62
Summary 63
References 64
INTRODUCTION
Microarray analysis of gene expression has become one of the most widely used
functional genomic tools, since its development in the mid 1990s (1,2). This is
reflected in the number of publications containing or arising from some aspect
of microarray technology. The development has allowed researchers from all
areas of biological research to simultaneously monitor the expression levels of
thousands of genes.