Replicate inconsistencies in qPCR - (Mar/10/2010 )

We currently started to have problems with replicates (triplicates) in some of our genes. There are inconsistencies between the replicates +/- 1.0 Ct, sometimes even more.
Cts are around 27-30 and the curve looks normal, sometimes some of the replicates have a different slope than the others. It's not that one of them is outliner, they all are.

What we tried to rule out ("fine" means there are no significant differences between replicates):
1) Pippeting error - you can't always rule it out, but we have an experienced technician, and there are reactions on same plate that are fine.
2) Machine error - other reactions on the same plate are fine, inconsistencies are not in the same position on the plate.
3) Mastermix problems - other reactions are fine, we use this mix from the start and get the best results with it.
4) Probe or primer problem - same primers and probe works fine on different template (we use UPL probes for the amplification, we have for some time and generaly with a good experience).
5) Template problem - other genes with the same template are fine and isolation procedure is still the same (column method from the mouse organs).

Anyone has an idea?

I experienced such inconsistencies when doing CHIP probably due to the nature of the template, but this is not the case. Also I would expect it in high Ct cycles like 35, but 27 seem quite low to me.

why do you want it to be 35? normally we don't consider anything around that a reliable result.

+/- 1 Ct is OK. The height also differs sometimes between replicates, no worries.

Curtis on Mar 10 2010, 05:13 PM said:

I meant I would expect high variability in high cycles, like 35, not around 27. But as you said, the results that high that are generaly not reliable.

well.. if its a quantitation... wont a 1 Ct difference matter a lot?? as it will be almost a difference of 3.3 folds??!!!
Specially if earlier it was not the case and now it is??!!!
What i mean is if a 1 Ct variation is obserevd all the time, it can be considered as an inherent variation but if its a new problem, then sumthing has to be wrong and i would think twice before using the values for quantitating anything!!
If its a yes no system, its still fine to go ahead.. depends on the variability that can be acceptable though!!
And i agree that the height differs.. but the threshold is at the log phase so the height does not matter wen analysing!!!
Correct me if i understand it wrong!!

yeah, a Ct difference of 1 suggests two-fold change, thats huge! .....If I were in your place I wouldnt accept this data. ......I was having similar problems when I was a newbie. Consider these points (that is, if you havent already): vortex and spin your master mix prep. really well and just before adding it to the plates. Because there is glycerol in the master mix solution, it tends to settle down. Secondly, each time you are adding the template, make sure that you change the tips, coz small volumes left behind in your tip can really give awry distribution of the sample. However, I am puzzled that you have started having this problem only recently!!!

Best

Check your ROX concentration and ensure you are measuring fluorescence in a linear range of you qPCR machine otherwise adjust the gain settings.
A good note about ROX is at http://www3.appliedbiosystems.com/cms/grou.../cms_053906.pdf

vladooo on Mar 22 2010, 04:21 PM said:

We use LightCycler 480, it doesn't use ROX reference.