ABSTRACT
Overview............................................................................................................... 243 Linear Normalization............................................................................................ 243 Intensity-Dependent Normalization ..................................................................... 244
Reference and Baseline Array.................................................................. 244 Smoothing Functions ................................................................................ 244 Quantile Normalization ............................................................................ 245 Qspline Normalization.............................................................................. 245 Z-Normalization........................................................................................ 245
Choosing a Normalization Method ...................................................................... 245 Normalization for Oligonucleotide-Based Arrays ............................................... 246 Performance of Normalization ............................................................................. 247 Data Transformation ............................................................................................. 247
Logarithmic Transformation..................................................................... 248 Power Transformation .............................................................................. 249 Variance Stabilizing Transformation (VST)............................................. 249 Generalized and Modified Logarithmic Transformations........................ 249
Conclusion ............................................................................................................ 251 References............................................................................................................. 251
With the sequencing of the human genome [1,2], newer generation microarrays are more capable than ever of presenting a global view of gene expression. Tens of thousands of data points can be obtained from a single assayed sample. With such a large payoff potential, experimental design and subsequent analysis methods are immensely important so that reliable results can be obtained from the data.
