Absolute quantification...

[Francisco]

Hello everyone. I was wondering why is relative quantification always applied in gene expression studies even though in many occasion standard curves are constructed. Is better to use Cq instead of concentration estimated through the standard curve?. If the correlation coefficient is good (bigger than 0.99) it should make no difference absolute or relative quantification. There are any problems if we normalize using absolute quantification approach?.

Thank you for your help.

Francisco C. Ceballos
Universidad Santiago de Compostela. Spain

[Mikael Kubista]

Francesco,
You are touching upon a very important fundamental point here. The term “Absolute quantification” as used in qPCR is very misleading. Using a standard curve to estimate concentrations of unknown samples is by no e unknown means an absolute quantitative approach. It is relative quantification of the unknown sample compared to the standards!

Good luck

[ayanndri]

hello everyone, I want to ask a basic question here. Can I determine the exact copy of the gene in the genome by absolute copy number determination?

[Mikael Kubista]

No. The name ”Absolute quantification” is misleading. Even if you estimate the copy number in your test sample by means of a standard curve, the estimate is based on comparison with the standard samples that were used to construct it. If the standards are in error, your estimate has a bias.

The closest you can get to absolute quantification is by using digital PCR: http://www.tataa.com/Services/Projects.html#Digital. But even here typically CNV is measured relative to another unamplified segment in the genome.

Good luck

[ayanndri]

Hello everyone, I am facing aproblem while constructing a standard curve. I am using serially 10 fold diluted PCR purified product as template for std curve preparation. The maximum conc. is 1.6 ng/microlit and diluted 6 times. I'm not getting Cp values accordingly. Should I dilute the template more?

Thanks

[ubali1]

Hello everybody...I am about to conduct an absolute quantification experiment for which I've already created a standard curve. Since I shall be reading multiple plates over the course of several days, I am curious to know if I need to repeat my standard curve, as well, with each run conducted on different days?

Thanks very much for your help!

Utsav Bali
UK

[Mikael Kubista]

It is not a good idea to perform separate standard curves on each plate/day. If you do that you will see a variation in efficiency, but this will not reflect true day-to-day variation, but rather depend on the particular noise you experience every time; unless, of course, you run very extensive standard curves (28 or standard points) each time. This is necessary if you introduce a change that can influence performance, such as new batch of primers, new master mix, new technician etc. If you don't expect systematic difference it is better to perform one single very accurate standard curve and use it to standardize all experiments. But you must run an inter-plate calibrator (IPC) to remove run-to-run variation. It is also advisable to run a couple of controls each time (there are recommended concentrations for those). I suggest using GenEx for the standard curve analysis, since it will also estimate the accuracy of the standard curve, identify any outliers, determine the dynamic range and, if you need, also LOD and LOQ. You can make your own IPC or you can get it from eg. TATAA Biocenter (www.tataa.com). You find also info on GenEx, which is from MultiD, on TATAAs website.

Good luck

[ubali1]

Thanks for the confirmation Mikael. I did think as much. I have been using GenEx for my standard curves and am quite pleased with it thus far. I shall most certainly consider running IPC's as recommended.

Cheers!
U. Bali

[ubali1]

Hello again Mikael,
We touched upon this issue very briefly during our discussion most recently at your qPCR workshop at the TUM. I just wanted to get your advice again on what the best methodology would be for pursuing an absolute qPCR experiment. Is it okay to reverse calibrate the unknown sample value and use total RNA as a loading control for data pre-processing, or should I run a housekeeping gene standard curve as well and use that as a loading control for my unknown samples?
Your advise would be much appreciated!

Utsav Bali
UK