Quality assessment of microarray experiments

Objectives: We describe a quality assurance procedure to maximize the value of oligonucleotide microarray expression profiles. Design, methods, and results: Background microarray noise was 82.2 ± 54.5 and 51.8 ± 12.4 units, respectively, before and after enacting the program (P < 0.0001). We also noted improved concordance of microarray expression fold-changes for selected genes with results of RT-PCR validation. Conclusions: This multi-step procedure, including quantification of RNA sample degradation and detection of outlier data points, has increased data quality from our microarray facility.