I am doing an experiment to check plasma microRNA expression profile in two time points in healthy individuals by using AB openarray protocol. In some samples the raw (time0) ct values are globally more than the ct of (time x) samples.
after normalization (to the Avg of ref genes) I had more than 2 fold increase in most of the expressed microRNAs which is not realistic and doesn't reflect the actual change in miR expression.
I have tried also normalization to mean, median but it gave me the same results.
Does anyone have any idea about the best normalization method in which we can remove such bias?
Hi Rana
Depending on the number of miRNAs that you are profiling, it would be nice to try the global normalization protocol. It works normally when you are profiling more than 20-30 miRNAs in each sample (typically screening panels)
By the way, what's your reference gene ?. There is not a clear and powerful normalization miRNA in plasma. It normally depends on your particular sample. I am currently using spike-in normalization.
Hope it helps. Best
Paco
I agree completely with what Paco said. Another thing to think about: since these samples were taken at two time points, they were also presumably processed at different times. It is possible that the global differences result from differing numbers of platelets/cells in one set of samples--from processing differences.
many thanks Francisco and Kenneth for your answers
for normalization I am using an average of miR-16 and miR-92a, but also I am trying to normalize for the mean or median and generally both methods gave similar results
Francisco is right. Did you incorporate spike in in your microRNA preparation? those RNA are introduced in the samples with known quantity and could be helpful for the normalization.
Best
- Badges
- Science topic
- Similar topics
- Biological Science
- Biochemistry
- Biomolecules
- Nucleic Acids
- RNA
- Small RNA
- microRNA