As analysis of miRNAs becomes increasingly popular, the resulting flood of microarray products out there for miRNA profiling is enough to make your head spin. So how do you decide which ones to use? If you’ve been stressing over that question, you’ll be happy to know that some good folks at Cambridge Research Institute and EMBL-EBI have already done much of the evaluation work for you.
The team utilized all three main methods of measuring miRNA expression: real-time reverse transcription PCR (qPCR), microarray hybridization and massively parallel/next-generation sequencing (NGS), each of them having well known benefits and drawbacks, to set up a study of six commercial miRNA microarray platforms with the results compared to both NGS and qPCR.
Putting miRNA Arrays Through the Ringer
The scientists tested each platform using RNA pools from normal breast tissue, a “normal-like” cell line PMC42, and a breast cancer line MCF7 using the manufacturer’s protocol. They rated the performance of each platform based on:
- Microarray hybridization: including signal-to-noise (SNR), CV between replicates and correlation between arrays.
- Probe mapping and specificity: They looked for proper miRNA targets and possible cross-hybridization issues.
- Correlation to NGS data: Comparing differential expressions (DE) between samples measured performance from each method.
- Correlation to qPCR results: Used to validate the DE results.
qPCR: The Gold Standard?
The investigators noticed during their experiments that qPCR results, considered the gold standard for miRNA expression and array validation, may not quite live up to the hype. They saw variable differential expression results by qPCR depending on the normalization used. This was especially true when looking at samples where overall miRNA levels vary (like cell-lines vs. tissues) and using a non-miRNA as a reference. To make sure you get the best results in your own experiments, the team recommends the use of several reference miRNAs as normalizers.
The Best miRNA Microarray Fits Your Needs
So which microarray platform was the best? Well, the short answer is: That depends. What they found was that the different arrays were suited to different experimental conditions, for a variety of reasons. Some might use more or less sample, others might have higher sensitivity but give more false positives, or yet others might be easier to handle making them better for high throughput situations. In other words, we can’t answer that question for you, but we can show you where to find the right info.
Find out which miRNA microarrays will work best for your own miRNA profiling experiments by checking out RNA, April 2010.