Although qRT-PCR has become a popular way to profile miRNAs, it has a dirty little secret that nobody likes to talk much about; data normalization. Without convenient “housekeeping genes” to use as internal references, like you would with mRNAs, researchers have had to settle for second rate options to normalize data from miRNA experiments. A group from Ghent University Hospital in Belgium has published a new method to make miRNA qRT-PCR data normalization much more effective.
By profiling hundreds of miRNAs in dozens of samples, the gents from Ghent showed how their new method, using the mean expression value of all miRNAs in a sample as the normalization factor, works better than what is currently used.
They also found that mean expression normalization:
- Lowered technical variation.
- Accurately found biological changes in patient samples and cell lines.
- Reduced “false positives”; under or over-estimated fold changes.
- Can identify miRNAs that “resemble the mean” for use as normalizers in smaller experiments.
Scope out this strategy and how it works at Genome Biology, June 2009