I am doing microRNA expression analysis, and we have used two technical replicates (platform QuantStudio 12k Flex). I know three technical replicates is the optimum but trying to balance technical and biological replicates, we decided to include two technical and four biological replicates. The thing is that we are calculating the mean for those technical replicates whose Ct values differs in no more than 0.3 but, what should we do when that difference is higher? Is there any other option than removing that result?
That's the problem about using only two technical replicates in your qPCR... I'm also doing it because I have quite a lot of samples.
What I usually do when the technical replicates differ:
Considering the fact that you have 4 biological replicates, you can look at the results of the other biol. replicates and see if they are similar to one of the values of the technical replicates that differ. It gives you a good estimate of what you would expext. If that's the case I would probably keep this one value.
Sometime it is pretty obvious when there is a problem with one of the replicates (you can see this sometimes in the melting curve as well).
However, if you are unsure about it and you have the feeling that something went wrong with the qPCR in the first place, I would definitely discard this one sample and redo it.
Dear colleague. Unfortunately, I can't answer Your question, because they do not do microRNA expression analysis
If your Cq values are in the Poisson noise range of the assay (e.g. >=35), you can accept a much higher degree of disagreement in the Cq values and arithmetically average those. But, if your Cq values are below ~34, identifying the problematic Cq values becomes somewhat easier (sometimes). Re-running problem samples is the better option though. I too run technical duplicates for large set-ups...
Thank you all for your comprehensive answers. They have been of great value
- Badges
- Science topic