Methods in Microarray Normalization PDF

Preface

Gene expression profiling using microarrays has been increasingly brought to task for the simple goal of ensuring that data from a particular tissue and a particular platform match the same tissue run on another platform in another lab. Expression arrays are becoming extremely valuable in the clinic and, as such, are being closely monitored by the Food and Drug Administration (FDA). MammaPrint uses the Agilent microarray system, which has been FDA approved. The National Institute for Standards has created the Microarray Quality Control Consortium in order to identify those factors that tend to throw lab-to-lab and cross-platform concordance off. They have made remarkable progress, and papers and new insight continue to pour out of their initial investment. The External RNA Consortium is purposed to create standards and spike-ins to allow proper measurements of array performance no matter what the location or environmental conditions are.

These efforts are moving the expression microarray from the experimental lab into the position of a trusted and widespread technological commodity. At that point it becomes a tool, but before that happens, the concept of data preprocessing and cross-platform concordance must be thoroughly addressed. Both topics have received much attention, but a line in the sand must be drawn beyond which we should assume that proper normalization will be used to minimize noise, bias, and cross-platform discordance, and to improve accuracy of biological interpretation.

Normalization and array preprocessing encompass a variety of data manipulations, but in this book we will look at those mathematical formulae that remove known and intrinsic biases in order to level the playing field—leaving expression data to stand on their own without accommodating the platform manufacturer, the fluorescent dye used, the varying quality of RNA, the laboratory, scanner, or any of a number of potential nuisance factors. In this book, many of the most respected authors in the field present their work for the reader to use and judge. In addition, several scientists have chosen to present their novel techniques for next-generation technologies such as the single nucleotide polymorphism (SNP) chip, the exon array, the comparative genomic hybridization (CGH) array, and the protein array.

Some useful references for the reader to engage in his or her own application of normalization include:

DNMAD (Diagnosis and Normalization for Microarray Data): http://dnmad. bioinfo.cnio.es/

Global normalization of two-color microarray data with loess: http://nbc11. biologie.uni-kl.de/framed/left/menu/auto/right/microarray/loess.shtml

SNOMAD: http://pevsnerlab.kennedykrieger.org/snomadinput.html Terry Speed's "Normalization of Microarray Data—How to Do It!": http:// www.maths.lth.se/bioinformatics/calendar/20010409/

Two-color DNA microarray normalization tool: http://128.122.133.135/cgi-bin/ rodrigo/normalization.cgi

University of Pittsburgh Department of Bioinformatics: http://bioinformatics. upmc.edu/index.html