Abstract

Microarray expression profiling is a complex experimental methodology comprising sequential steps: fluorescent or radioactive labelling of extracted RNA, hybridisation to the array, washing, scanning, and image processing of the array and data analysis. Each of these steps is a possible source of artefacts that potentially can influence the final result, and hence the evaluation of differential gene expression. In addition, the quality of expression profiling experimental data also depends upon the design and quality of the arrays themselves. Here we review specific problems and pitfalls that may occur during expression profiling experiments and suggest possible remedies.

Keywords: expression profiling, microarray artefacts, dna-chip, hybridisation, fluorescent labelling, image processing, cross-hybridisation, hybridisation specificity