4 replies to this topic

[jiang21]

Hi there,
Is anyone using plasmid to make standard curves in the real time? I am kind of confused. I followed the protocol as other people suggested.
I made the plasmids for target gene and 18s. I used them as standards. But since I have a lot of samples, I did multiple plates. For each plate, I did standard curves and a reference sample.  I used bio-rad machine, so I got Ct values and Log SQ.

My question is - how to adjust values between plates? How to use 18s in this case? I am still confusing. Is this method absolute or relative quantification...

Thanks a lot,

[jiang21]

Can anyone give me a reply??

[Trof]

Depends on the method of quantification you use. You have to decide if you want to know the absolute amount of your target gene or relative ratio compared to the calibrator sample (also called reference sample). In absolute quantification you know the copy number of your standards and can calculate the copy number of your samples, you use standard curve method (you can also have a relative standard curve method, where you don't know the exact copy of standards, but you know how they are diluted). In relative quantification you usualy need the standards only to assess eficiency for calculation of eficiency corrected ratio between your calibrator and other samples (Pfaffl method). All methods are described here.
You can make relative ratio from the absolute quantification data, if you divide it accordingly, but I find using Pfaffl method easier, you only need the Cts and efficiency. The machine software should calculate efficiency or you can do it in Excel, by plotting the standards on a logarithmic scale of concentration and Ct (linear). Then you make a linear regression and get a slope. Efficiency is 10^-1/slope, should be close to 2.


You need to have the calibrator sample always on the plate, since you normalise to it. As I understand you have. If you use Pfaffl you don't actually need the standards on every plate, if you use the standard curve method, you do. You may use all the standard curves to calculate efficiency and then average it, to make it more accurate.

Just always normalise all samples to the calibrator on the same plate, the calibrator is always 1. Then you can compare samples between plates, because you used the same calibrator.

[jiang21]

Thanks a lot for your reply, very helpful.  

I still got a question. If I calculate into copy number, do I still need a housekeeping gene to adjust the copy numbers, because there may be differences in regards to reverse transcription efficiency of all samples or even pipette errors during dilution of RT products.

[Trof]

jiang21, on Jul 7 2010, 06:22 AM, said:

I still got a question. If I calculate into copy number, do I still need a housekeeping gene to adjust the copy numbers, because there may be differences in regards to reverse transcription efficiency of all samples or even pipette errors during dilution of RT products.

Yes, even absolute quantification can be normalised to a housekeeping gene. You adjust your gene of interest copy number for the same amount of housekeeping gene, for example 1000 copies.