Hi everyone,
I'm looking for a little help in the analysis of my qPCR raw results. I hope you'll be able to help me. Thanks in advance.
Here is my example:
I have 3 genes: 1 target gene, and 2 housekeeping genes.
I have estimated the PCR efficiency for these genes and their associated primers: E(target gene) = 2.1, E(HK gene 1) = 1.9, E(HK gene 2) = 2.2
I have Ct raw results for 2 technical replicates each time
I have a Treatment condition, and a Control condition, for all of the 3 genes (target gene and the 2 housekeeping genes)
Finally I have these data for 3 biological replicates
(see the attach Excel document).
From these data I have calculated the normalized expression of my target gene (using the control condition as a calibrator, and the housekeeping genes for final normalization). All the calculations are in the Excel document.
In the end I have a mean of the normalized expression for my target gene, which is great. However, I don't know yet how to obtain the standard error for this mean. I can calculate the standard error from the 3 normalized expressions from my 3 biological replicate (SE of 1.39, 1.55 and 1.26, which is easy in R or Excel), but from I what I read on internet (but I didn't understand it yet) there's apparently a way to calculate the SE value from the different SE obtained all along the calculation process (SE from the mean of the technical replicate values of CT, then SE from the relative quantity calculation, and finally SE from the normalization process).
Does anyone is familiar with that and could provide a little help or at least a little look at the attached document to confirm that my calculations are correct. From then you'll probably be able to lead me towards the correct calculations for my SE value.
Thanks a lot in advance
Regards
Marc
Hello Marc, this is such a common problem. I don't pretend to be an authority on statistical analysis or reporting of error in qPCR, but I did find this spreadsheet to be helpful to derive and propagate the SE. It is a modified version of the error calculated in GeNorm/qBase which seems to be the "gold standard" of qPCR statistics (at least as far as I've seen). I think I found this spreadsheet somewhere here on ResearchGate under a similar question.
This spreadsheet has 3 different tabs; the first two are from the author (Amy Replogle) and only incorporate two replicates. I created the third sheet to allow for three replicates, and I think all the excel formulae are correct, but if you choose to use it, please check the equations thoroughly.
I don't really like using this approach, because I often run my different biological reps on separate PCR plates, and their Ct values over different runs can vary quite a lot (even though the dCt between a treatment and control, or a target gene and reference gene remains the same). Subsequently, I have a large measure of error that starts with the very first calculation. I prefer to test the different genes and treatment/controls within the same plate. I think this is more accurate, and I don't see why I can't calculate the relative expression of a single biorep without errors, then average (and measure error) across the different bioreps. But, as I said, I am not an expert and this attitude clearly goes against what VanDeSompele and GeNorm recommend.
Best of luck to you.
Thank you Amy i'll have a look at it ASAP.
I agree with you on the SE thing which, in my mind also, has to be calculated from the expression values obtained for each of the biological replicates. just a classical SE calculation, relevant of the biological diversity, and not a SE based on the qPCR technique itself.
I've been in touch with other researchers since my last message. If I get anything new I'll post here some explanations of what I've been doing and what solutions I found.
Thanks again
- Badges
- Science method