acceptable variation for biological replicate - (Mar/28/2012 )

Hi all,

I am just wondering of what would be the acceptable level of variation for biological replicate in real time PCR. I did my statistic on 'fold change' which is derived from the delta delta ct for about 18 genes per treatment group. Some of my gene have low variation and some have high CV (greater than 20%). is that normal? The ct values for the housekeeping gene itself was not significantly different across treatment.

What is the variation of your replicates for target genes?
Usually higher Cts have higher variation, that is expectable. This variation projects into the variation of normalised result.
Since housekeeping genes are generally high expressed, I would suspect higher Ct's or just higher variation in some of the target genes. Some applications (REST I think) have a cut-off for SD of Ct replicates 0.2 what is bigger is too variable.
Sometimes when you have more than duplicates, it can help to look at the curves, find the outlier and remove it from calculation to decrease the SD, but you can't do it if the curves are evenly distributed (which those variations caused by later Ct's are, since there some small number problems comes to effect), one would need to be outlying for different reason, pipetting or other error.

Thanks Trof!