Hi Eric,

To use only one miR16 for normalization for qPCR is not enough. It usually recommeded to have at least 5-6 miRs for normalization. In my opinion miR16 is not perfect choice, because it is rather a plasma biomarker. In your case you deal with cancer tissue of epithelial origin, so you need some epithelial markers or the circulating miRs from biofluids of normal breast (breast milk or colostrum). To some extent validated miRs for breast you find in the attached study. They are forensic researchers and have a deal with multiple types of normal tissues.   

Good luck!

Thank you for the answer and the attached paper, Andrey!

Pardon the amateur technical question, but how does one mathematically compute multiple miRs as control/housekeeping genes using the Delta-Delta-Ct method? Do you simply take the the mean of all normalizer Ct values for the two groups and use that for the calculation? I.e. Say you picked 5 housekeeping miRs and had groups of 5 versus 3. In this scenario you would average the 5 miRs (in triplicate for technical replicates) for the normal versus diseased group and use that "aggregate" Ct to calculate the deltas against your miR gene of interest? Also, can you include an exogenous spike-in control miR to this normalizing group and calculation or should that be separate? Any technical or statistical details would be appreciated.

Thanks again!

-eric  

In this scenario you would average the 5 miRs (in triplicate for technical replicates) for the normal versus diseased group and use that "aggregate" Ct to calculate the deltas against your miR gene of interest?

Exactly: "cocktail" of housekeeping miRs versus miRs of interest in normal tissue versus tumor. Prior, I would recommend to evaluate expression of housekeeping miRs in normal tissue only in order to set up a treshold for futher analyzes in tumor. 

Thanks again! Using this approach for just the pooled miR-16 and cel-miR-39 control Ct values by Delta-delta-Ct yielded VERY close fold changes to what our deep-sequencing came up with. I will continue looking in to candidate controls, but your advice has been very helpful!

Hey I'm sure this experiment is probably finished by now but it's important to note that a spike-in-control can not be merged with a housekeeper. A spike-in-control, is when you pipette a known volume in during your cDNA synthesis step. It's there for the purpose a reaffirming your cDNA synthesis step was done correctly. The problem with this can be seen by the example below:

You dilute your RNA sample 1 to 20ng/ul and add your dNTPs, mastermix, buffer, enzyme, H20 and spike-in.

After your PCR your results could look like this:

miR of interest = CT of 28

Housekeeper 1 = CT of 24

Housekeeper 2 = CT of 20

Spike-in = CT of 20

Average of Housekeeper 1 + Housekeeper 2 = CT of 22

Average of Housekeeper 1 + Spike-in = CT of 22

So this has produced identical values.

The problem comes for when you dilute your next sample and you end up diluting it to 20.2ng/ul and add your dNTPs, mastermix, buffer, enzyme, H20 and spike-in.

After your PCR your results could look like this:

miR of interest = CT of 27

Housekeeper 1 = CT of 23.5 (slightly lower)

Housekeeper 2 = CT of 19.5 (slightly lower)

Spike-in = CT of 20 (remains the same)

Average of Housekeeper 1 + Housekeeper 2 = CT of 21.5

Average of Housekeeper 1 + spike-in = CT of 21

There's now only half of a CT difference between sample 1 and sample 2 in the average of the two housekeepers.

However as the spike-in is a known concentration it will still generate the same CT value as in the first sample and now you're starting to see a large CT difference from a small pipetting error between sample 1 and sample 2.

The point being that your housekeepers are endogenous values that are altered slightly along with your gene of interest when your pipetting is altered slightly throughout any interaction with your sample, i.e. Extraction, dilution, cDNA and RT-qPCR.

Your housekeeper is a known value, that you have put in that neither gives you any information about the endogenous profile of the sample, nor does it help with housekeepers, it merely confirms that with a consistent CT of 20 (or whatever consistent value you end up with), that your samples haven't degraded (from freeze thawing) or you haven't pipetted incorrectly, or you haven't miscalculated.

Hope this helps.