Dear Siviseth.

Expression is a very flexible and unstable, dynamic parameter. Did you use the same sample in both experiments? Or only the same type of sample (other aliquot of spheroid cells)? How many folds you mean when said "the pattern of expression decreases"?

Frencly speaking the best way to repeat an experiment is to eliminate all the differences between two series. But I am sure thet you have already know it ).

My expirience tell me that 10-fold (+/- 2 fold) is ok to see the differences. More precise in the case of expression profiling - difficult to repeat.

You probably need to provide a lot more detail.

How are you isolating your RNA and preparing cDNA? miRs are tricky things.

How many biological replicates are you performing? How many technical replicates?

What reference genes are you using? Other miRs? If so, how many and how do you know they're valid?

How are you comparing expression between your two conditions? "Increases vs decreases" implies a temporal element: are you looking at changes over time?

How are you 'running your samples'? What assay, what conditions?

Dear Siviseth.

Expression is a very flexible and unstable, dynamic parameter. Did you use the same sample in both experiments? Or only the same type of sample (other aliquot of spheroid cells)? How many folds you mean when said "the pattern of expression decreases"?

Frencly speaking the best way to repeat an experiment is to eliminate all the differences between two series. But I am sure thet you have already know it ).

My expirience tell me that 10-fold (+/- 2 fold) is ok to see the differences. More precise in the case of expression profiling - difficult to repeat.

Most likely your experimental setting and/or protocol is not reliable. This can (among other things) depend on the experimentator, the apparatures, storage of sample, the used technique or chemicals/primers/first strandsynthesis etc.

John Hildyard Thanks for spending your time answering my question.

I extracted my RNA withTRIZOL and synthesized my cDNA with SuperScriptIII First-strand Synthesis system. I replicated each sample. And for reference gene, I used GAPDH. I run my sample with SYBR.

Another question is how can we know that the primers we designed are valid or not before making an order.

Andrei S. Babenka Thanks for taking your time giving me some good advice.

I used the same samples for my experiment. There was no such a big change (like you mentioned - 10 folds), I just saw the pattern is increasing or decreasing.

Andreas Eisenreich Thanks for taking your time answering my question.

I have thought of those possibilities that could lead to an error of my experiment too, but I have not figured out which one yet.

I wish you good luck for further experiments.

Hello, Sokviseth. First of all, in our experience the use of only one reference gene in unacceptable. Second, GAPDH is very bad as a reference gene since it has pseudogenes. Moreover, there is plenty of reference miRNAs like U6, we also used SNORD48, SNORD47, and SNORD44. And the last, but not the least, you should test your supposed reference miRNAs in your experimental conditions to be sure that their levels are stable through out all your experimental and control groups.

you run the experiment from new cDNA or stored cDNA of previous experiment.

Rakesh Kumar The first experiment would be new cDNA as cDNA is newly synthesized. Later on, this cDNA is stored and is used again in the next next experiments.

Hello Moeng,

I would like to know something.

Which method did you use for miRNAs expression?

In my case, I used stem-loop RT PCR and let me know how did you design GAPDH as a normalizer for miRNAs if you used stem-loop RT primer.

stored cDNA would give different pattern of expression, It depends on storage condition and storing solution.