quantitative methylation specific PCR analysis

[Liesbeth]

Hello,

For my thesis, I did quantitative methylation specific PCR (qMSP) on colon cancer cell lines for CpG island containing regions in three different genes. And this for different conditions.

Now, I was trying to analyse this data.
I was wondering if you can use the same analysis for qMSP as for gene expression (so, with the exponential calculations)?
I read in an article that you can use the following calculation: average Ct-value of triplicates of gene of interest/average Ct-value of triplicates of reference gene x 1000.
So dividing by this reference gene value is enough to normalise?
And what about error bar calculations?

Thanks in advance.

Liesbeth
2nd master Biochemistry & Biotechnology
Ghent University, Belgium

[Mikael Kubista]

Dear Liesbeth,

It is correct that normalization can be written as:

(arithmetic) average of Cq's of the gene of interest divided with the (arithmetic) average of the Cq's of reference gene(-s).

High delta(Cq) will correspond to low expression and vice versa, which may be impractical for plotting (although it does not affect statistics). To reverse the signal, in GenEx delta(Cq) values are converted to relative quantities, and then back to log scale again by taking the log base 2.

Rule is that technical replicates shall be averaged, while statistical analysis, including the plotting of error bars, shall be performed on biological replicates. Biological replicates in general refer to different subjects, which typically are different individuals. The reason is that the statistical result should reflect differences between populations comparing the treatment effect with biological variability. When studying cell cultures things situation is different, since there is no biological variability (unless cultures of several strains are compared with and without treatment). Hence, there is no rule what measurements on cell lines, if any, should be analyzed using statistics. Whatever you chose, it will strictly not be a comparison of biological variability with treatment effect.

Good luck