by mchlbrmn » Aug 02 2012 12:29 am
Hello. I will first say that I am not a qPCR expert, but I do use it sometimes and these are my impressions.
If you want to know if your Cp's over 30 are valid, you should run a series of 1/10 dilutions (in duplicate or triplicate) with a strong positive sample (relative to your other samples). If the efficiency is a perfect 2fold/cycle, the spacing should be 3.3 cycles. It may be less in the concentrated samples if inhibition is delaying the concentrated sample. It may be more in the dilute sample if the PCR is starting to fail. At very high Ct it may become erratic because initial target DNA molecule numbers are so low that random segregation of the molecules can put more copies in one tube than another. Looking at the dilution series you can judge what range of Ct's is working well (for this template at least, but one can't always do everything) and call the Ct's greater than this 'beyond the range of detetection in this experiment', and those Ct's lower than the good range are also not good. You may not achieve a perfect 3.32 in every dilution, but it should not be too far away.
Looking back, I also noticed that there may have been miscommunication involving the difference between a "single copy gene" (=one copy/genome, but many in the initial tube) and a "single copy of a gene" (=one target molecule of a gene present in the tube).
Oh, I just noticed that you listed PCR efficiency values. So, you already did the dilution curves, and you should be able to look at these, if they extend far enough, to see what Cp value your experiment should be valid.
Another thing to consider is if (when using SYBR PCR) the melt curve gives a single peak. If high Cp samples show an extra peak(s) appearing, and the Cp spacing between 1/10 dilutions becomes lower, this indicates that a nonspecific PCR product is appearing, and affecting the Cp.
If a part of the dilution standard curve is not good, when you remove these samples and recalculate the efficiency, it should become better. (So, instead of having Cp's of 26 and 35 for two samples, the Cp's may now be 26 for one sample and greater than 32, or not detectable within the range of the experiment for the other).
Actually, the efficiency does not need to be 2.0 (and the Cp spacing 3.2) as long as it stays consistent (for example 3.0 Cp between every dilution). Your efficiencies listed look OK to me, but I wonder why there are none that are 1.9 or 2.0. Perhaps if the Cp range of the experiment is reduced, as I described, efficiency would improve (maybe in the last dilution the PCR is starting to fail). However, as long as the Cp's are still increasing with the dilutions, even if the spacing is not perfect, the experiment is still quantitative to some extent (I don't know how much), especially if you are looking at very major differences in expression.