Relative RT-PCR, multiplate and calibration - (Mar/03/2010 )

I would like to ask your opinion on the following:

i have 60 brain samples to run in RT-PCR. I´m using GAPDH as a housekeeping gene. I run all the samples in triplets and put calibration sample and NTC to every plate.

Because I have the calibration sample in every plate, can i run full plates of only target gene and full plates of only GAPDH, calibrate target gene plates and GAPDH plates to match and then calculate the relative differences??

Usually it is adviced to put the target gene samples and endogenous control samples on a same plate, but as we are talking about relative changes and plates can be calibrated within the group, it should be ok, no?

Thanks for your help!!

Hi, yes this is possible. But you just have to calibrate within your target gene plates and within your GAPDH plates, not between target and GAPDH. Inter-run-calibrators are used to correct for technical variances between different runs but this doesn't concern you for the reference gene normalisation.

One hint: run the calibrator sample at least in quadruplicates on the plates.

tea-test on Mar 3 2010, 07:47 PM said:

Hi!

A follow-up question once the plates are ran. Or just to be sure.

I use ABI 7300 system where I can put all the target gene plates and all the GAPDH plates (separately) to a plate study and adjust a same threshold to them for the Ct calculation. Do I still have to calibrate the Ct values in excel by using the calibrator samples ran in every plate? I would think so, no?

I´ve asked the ABI helpdesc but I don´t get a clear answer; I think we are talking about different things...

Thank you again!

Hi,

I don't have experience with the ABI7300 system, I'm working on a 7900 and StepOne plus. If you run the plates under "RQ" mode on the 7900 you can analyze them as a study using the RQ Manager software (which comes together with the SDS software) with the drawbacks that you are limited to only 10 plates and it cannot normalize to multiple reference genes.
If you are using the ddCt method for the analysis you can also use the free DataAssist software from ABI DataAssist2.0.
There you can upload an virtually unlimited number of plates, it can handle multiple reference genes (and can even determine the most stable ones) and it can also calculate the p-value between biological groups. Drawback: I don't know how to save an analyzed study. So if you want to change something you have to adjust everything again which can be nasty if you have many different biological groups. Maybe it's easier if you upload a sample design file.


If you got already your RQ values I would say you don't have to calibrate again in excel because that is the idea behind combining different plates to a study, the software should automatically calibrate to the respective calibrator on the same plate during analysis.

Hello once again,

when I have now 7 plates of target gene and 7 plates of housekeeping gene do I just simply calibrate all other 6 plates in relation to the first one (for example), or to say: all 6 others to the one that I select as a "reference plate". Too much thinking about this, getting unsure about my own thinking.

Thanks again!!!