How do you choose the threshold?

3 posts / 0 new
Last post
RosieBS's picture
How do you choose the threshold?

 I am trying to use QPCR to determine DNA copy number of a transgene as it is suspected to have dropped within a colony. So far I've used the comparative method of quantification rather than absolute method. I'm carrying out duplex reactions on an ABI 7500 system. My control sample has been extracted using a different method to my test samples. 

I am having difficulty working out how to analyse my data as the level at which I set the threshold is having a massive impact on the results. The baseline setting also seems to have a large impact on the results, even when it is set within the apparent baseline and even when only small changes are made to the settings. 

I understand that I should try and set the threshold to the lowest possible point during the exponential phase. This varies a lot between samples. I've tried setting it according to this rule but even slight alterations seem to have big implications in the results. I think this might have something to do with the DNAs being extracted using different methods and the effect it has on the PCR efficiency. I've looked up methods of calculating efficiency and standardising for this, but these all seem to refer back to the Ct, which of course depends on the threshold I have set. 

Any help on this would be really appreciated. I feel like I'm going round in circles! 

Ivan Delgado
Ivan Delgado's picture

It is close to impossible to give you any advice with so little information about what you are doing and what your results are. If you can share more about what you are doing (the efficiencies of your assays for example), and the results you are getting (are your amplification curves sigmoidal, your Ct values within an acceptable range between your assays, etc) we may be able to share some insights. 

As it is, extracting your controls and your samples using different techniques can cause the kinds of problems you are having.

RomeoLima's picture
Hi Rosie

Hi Rosie

If your are looking for other methods to calculate the efficiency , I would suggest to take a look of the LinReg program or the linear regression efficiency method (LRE). This two methods allow you to estimate your amplification efficiency by using  the fluorescence reading per sample (you do not need to construct a standard curve).  There is a JavaScript program for the LRE,and you can find the LinReg on the web too.