Statistics and Data Analysis for Microarrays Using R and Bioconductor, 2nd Edition

Chapter 20

Data preprocessing and

normalization

If at firs t you don’t succeed, transform your data set.

—Unknown [165]

20.1 Introduction

Preprocessing is a step that extracts or enhances meaningful data characteris-

tics. Sometimes, preprocessing prepares the data fo r the application of certain

data ana lysis metho ds. A typical example of preprocessing is taking the log-

arithm of the raw value s. Normalization is a particular type of preprocessing

done to account for systematic differences across data sets. A typical example

of normalization is modifying the values in order to compensate for the dif-

ferent dye efficiency in the two channel microarray