Real Time PCR - Relative Quantitation - Choice of calibrator and calculations (Jul/08/2005 )
Hello all.
I am using SYBR green relative quantitation to compare the relative transcript abundance of a gene in fibers of two cotton cultivar over 4 developmental time-points. Our endogenous reference for normalization is a ubiquiton-conjugating protein. We can look at normalized relative transcript abundance (2^-DCt) and determine fold changes in transcript in one cultivar relative to the other. My question is in regards to the use of a calibrator to determine (2^DDCt). We are doing both cultivars for one time-point per plate. This includes three experimental replications and three biological replications. Amplification efficiency checks out fine for all primer pairs. Can you choose a calibrator from one time-point (say the lowest transcript regardless of cultivar - or the largest DCt value after subracting the normalizer) and then apply this to all DCt values for all timepoints and cultivars to calculate DDCt? It just seems strange to choose a calibrator from one plate (time-point) and then apply this to all other plates (time-points) without the calibrator present on those plates. It seems that the calibrator should be selected and present in all runs. For example, select one cultivar in the ealiest developmental time-point for the calibrator and include these samples on each 96-well plate to calibrate each plate (time-point) independent of the others. Since this calibrator is always present, you can then make relative transcript abundance comparisons between cultivars across all time-points. WHEW!! Maybe I am over-thinking the concept of relative quantitation, but I realize the value of proper statistics and experimental design. Thanks for any advice.
Cheers,
Dr. Doug Hinchliffe
Can you choose a calibrator from one time-point (say the lowest transcript regardless of cultivar - or the largest DCt value after subracting the normalizer) and then apply this to all DCt values for all timepoints and cultivars to calculate DDCt? It just seems strange to choose a calibrator from one plate (time-point) and then apply this to all other plates (time-points) without the calibrator present on those plates. It seems that the calibrator should be selected and present in all runs. For example, select one cultivar in the ealiest developmental time-point for the calibrator and include these samples on each 96-well plate to calibrate each plate (time-point) independent of the others. Since this calibrator is always present, you can then make relative transcript abundance comparisons between cultivars across all time-points.
I think you should include the calibrator for each and every run. Mostly my calibrators come out at approx the same Ct between runs, but there have been instances when the machines acts up - so the calibrator comes out at a different Ct (if you didn't include the calibrator, you would assume that the Ct change is a real change for your samples). So including the calibrator allows you to recognise this and exclude/repeat that run.
Of course, if the calibrator Ct changes but your machine is fine, then that suggests some cDNA degradation?
Also, I understand the bit about the sample replicates, but when you say biological replicates, you are refering to replicate expts done on the cultivar at a single time point right? In that case, for each expt, the Ct of the sample for each expt would have to be normalised with their own calibrator.
No doubt this may seem a waste of resources if Ct for calibrator comes out the same with each expt, but actually doing it makes the results more reliable.
Let me know if I got something wrong