Microarray has revolutionized the study of molecular biology, especially the application in clinical diagnostics. When used in clinical diagnostics, microarray has to meet a high degree of reproducibility, reliability and quality in order to become a standard tool. Repeatability and reproducibility are essential for providing the best data and process control. The real challenge for microarray is, however, how to produce consistent and reliable data. The variance of microarray data is contributed by the quality of sample source, the quality of glass substrates, hybridization, and probe labeling and spotting. The quality of coated glass substrates is one of the main factors. This paper is focusing on discussing how to optimize coating conditions to improve the slide quality, consequently improve the data quality such as sensitivity and data reliability.
Efficient hybridization of complementary strands of DNA is the underlying principle of all microarray-based techniques for gene expression analysis. Recently studies have been published to assess oligonucleotide (55-70 bases) performance on glass-slide microarrays and stress advantages over the cDNA arrays. Importantly the oligo arrays eliminate possible failure in PCR amplifications and attain sequence optimization. In the present study, we have used 60mer oligo microarrays to investigate the effect of target (immobilized on the glass slides) and probe concentrations and possible probe interactions on hybridization. Scanner calibration slides (manufactured by Full Moon BioSystems) were used to concert the fluorescence signals into fluorophore per μm<sup>2</sup> in order to eliminate possible variation from scanner. The retention of the target was determined based on mock hybridization using Cy3-labeled oligonucleotide. We found that hybridization signals fell within the linear response range when the target concentration (printing solution) was equal or less than 2.5 μM. With fixed target concentrations, there is a non-linear relationship between the probe concentration and the hybridization signal. Dual-probe hybridization measurements suggest that hybridization of probes is not ideally independent. In this study, hybridization signal from Cy5 is consistently lower than that from Cy3.