Summary
Objectives:
In this paper we give an overview of post-hybridization quality control methods for gene expression chips, including methods for the gene/spot level, the hybridization/chip level and the process level. We present quality control methods that can be applied after hybridization and image analysis, i.e. that use data from the chip experiment itself. Wet lab quality control steps, which should be applied before the probe is measured on a chip, are not discussed. This review is aimed towards statisticians and data analysts.
Methods:
We give examples of some of the quality control measures available for spotted cDNA and Affymetrix GeneChips®, the most common chip types. As quality control measures are technology and design-dependent, we will stress on methods that have the potential to be applied platform-independently.
Results:
Quality control should identify poor quality chips or hybridizations, as well as faulty measurements for individual genes/spots. Additionally, high throughput laboratories processing several tens or hundreds of microarrays per week have the need for an appropriate process control to be able to identify changes in the production process as early as possible.
Conclusion:
Microarrays have become a standard research tool for biologists and medical researchers. As a consequence, there is a great need for standardized quality control, as false findings due to problem in data quality can lead to a substantial loss of resources.
Quantitative quality control in microarray experiments and the application in data filtering, normalization and false positive rate prediction