1. Introduction
miRNAs are a set of naturally-occurring small single-stranded non-coding RNA molecules. They regulate more than 30% of all human genes and mediate gene expression at the post-transcriptional and translational levels in both plants and animals. Many miRNAs are evolutionarily conserved and are involved in many essential biological processes, such as development, cell growth, differentiation, apoptosis and tumorigenesis [1,2,3,4,5,6,7]. The substantial value of miRNAs has been reported for diagnostic and prognostic determination, as well as for eventual therapeutic intervention [8,9]. Although the first miRNA, lin-4, was initially discovered over a decade ago, the significance of miRNA research has not been appreciated until recently with the discoveries of thousands of miRNAs in worm, fly and mammalian genomes [10,11,12,13].
Gene expression microarray technology is a very powerful high throughput tool capable of monitoring the expression of thousands of genes in an organism simultaneously [14]. Probe-target hybridization is the central concept to determine the relative abundance of nucleic acid sequences through fluorescence-based detection [15]. In the past few decades, microarrays have been increasingly utilized to investigate the complex molecular interactions in biological systems. Along with increasing interest in miRNAs, most well-established molecular and biological technologies have been successfully transferred into miRNA research, including the microarray and quantitative real-time polymerase chain reaction (qRT-PCR). Currently, there are many commercial miRNA microarray platforms available, including the products from Agilent Technologies, Ambion Inc., Exiqon, Invitrogen, etc.
Like most analytical platforms, errors are introduced to miRNA microarrays in almost every single step in the process of acquiring and analyzing microarray data, including sample preparation, sample storage, dying, hybridization, scanning, image processing and equipment errors, among many others. In addition, due to miRNA’s unique signatures, such as small total number and low expression in the majority, the transplantation of traditional normalization methods for mRNA/cDNA profiling analysis is not able to provide a suitable solution for miRNA profiling analysis. The emphasis of this study will be placed on the data analysis techniques applied to the final miRNA expression data profiled with three different platforms: two versions of miRCURY LNA microRNA arrays and one Luminex FlexmiR microRNA Human Panel. The results are validated with the qRT-PCR results by TaqMan Array Human microRNA Panel.