DNA Methylation Microarrays and Quality Control
A microarray experiment involves various steps, each of which is prone to error. For random error, repeated measurements will improve the precision. For systematic error, data normalization is to remove the bias. In either case, we need to have control of the quality of microarray measurements so that the results are reliable and credible. To this end, this chapter reviews the principles of gene expression microarrays which, after adaption, enable genome-wide DNA methylation profiling. We introduce two-color microarrays with probes from CpG (cytosine-guanine) island libraries and one-color oligonucleotide tiling arrays with probes covering nonrepetitive portions of the genome region of interest. Two popular enrichments of DNA methylation fragments for microarray hybridization are introduced; one is based on methylation sensitive restriction enzymes and the other on immunoprecipitation with antibodies against methylated DNA.