Normalization to Standard Gene: what does this mean - (Nov/30/2012 )

Dear Guys,
I all new in reverse transcription-PCR, I need ur advice.
I read that in reverse transcription-PCR results, the test with low Ct values this means the presence of starting material then, it is more positive.
but when I saw the papers, I can see that they express their results related to standard gene like GADH.
what was in my mind:
I will divide my Ct value of test to Ct value of STD gene,
the I get a number, I will use this number and put it in my graph.
My first question, Does this is ok??
but when I did this, I got one test has value of 2
and other test has value of 1.3
So I guess this means that first one is expressing my gene more that the second one.
but when I think that higher values means less expression as I am using Ct value?? I get confused.
So please help
Second issue.
how I can get Standard deviation (STD) or Standard error (STE) for this results?
Divide STD from tested gene samples to STD of standard gene??
or what I can do
Thanks Guys for ur help

Guys I meant Real time PCR, but I do not know why It is written as reverse transcription PCR.
I guess from my previous post and this post, I am getting totally mad

Have a look for the Pfaffl method or papers and techniques on delta delta Ct.

Dear

bob1



Thank u for ur priceless advice.
I have downloaded some papers to discuss the use of Pfaffl method and ddCt.
But I have some questiion, I could not understand.
I wonder if u could help me to understand these points.
At first, ddCt method:
after several equation u would reach a result said
amount of gene expressed = 2^{delta (detal Cttest - delta Ct standard)}

=2^{-delta<(Cttest gene-CTreference Gene) - (Ct standard test gene - CT reference gene).}
my question ??
If I am using testing the expression of albumin for example.
and my refernece gene is B-actin.
and I used standard Albumin plasmide and made serial dilution of my standard
Starting from 10 ng.
so which Ct value I should use for my STD as I would have different Ct values for each of my STD serial dilution.
Shall I use CT value for STD that is near the CT value of my tested gene??.
Regarding statistical analysis, how I can compare between results??
Shall I compare it to my reference gene, or my STD tested gene expression (I think I should compare it to reference).
then Standard deviation or Standard error should be calculated form the obtained Ct values of tested gene solely???
or how can get a represetative Standard deviation after these mathematical equations.

Regarding Pfaffl method.
At the end it assumed that ratio of expressed gene = E ^{deltaCT
=(Ct control-Ct Sample)}

Regarding Pfaffl method. ^{E=(Ct control-Ct Sample)}
At the end it assumed that ratio of expressed gene = E ^{deltaCT
so = E=(Ct standard - Ct test).}
This method depend on that Ct control - Ct of reference gene will be the same.
How this is,
I cant understand forgive me.
How CT of B-Actin will be equivalent to albumin (in my example).
or I must choose the control dilution which fulfil this condition.
and use this CT of this dilution to calculate ration in the above equation.
if not, so why I use reference gene from beginning ?? if it does not enter the equation of normalization?

Regarding efficiency (E) I would use the efficiency calculated from the curve of Ct against log serial dilution of STD of my tested gene?
Thanks in advance for ur help
Madelin

I think (it has been quite a while since I last did this) that for the reference gene you would take a single point and use that as the basis for all the samples. If you change the reference gene value, then your calculated value will change too.

The efficiency is calculated off the slope of the standard curve. It should be about -3 for an ideal slope.

There are lots of people on here who will be much better qualified to answer these questions than I am, so I'll leave it to them ...

madelingirly: I just got lost in what you are actually asking, so I will just go from beginning.

The efficiency (E) in the Pffafl equation can be calculated from any sample, but instead of using a dilution of a plasmid I would recommend doing a dilution of some highly possitive sample (for both actin and albumin). Diluted plasmids (especially without any "dummy" nucleic acid present in the dilution) are very noncomplex templates and their efficiency may be different from the efficiency of a real cDNA sample.

For dilution curve you need at least 3-4 points (dilutions), each in triplicates preferably, but the dilutions are not needed to be 10 times if Ct is too high even though you choosed the most abundant sample (you should calculate that every 10-fold dilution will increase the Ct around 3, and values after some Ct 30-33 are hard to quatify and very variable), but only 5 times for example. You need to run standard curve only once, and you can use the efficiency values for all subsequent runs with same primers/(probe). You don't work with any standards from now on, just compare cDNA samples. (and for delta-delta method, you don't need to calculate efficiency at all, but sou should know it since it can tell you if your assay is good or not)

Now you have to measure both reference gene and the target one in all samples. One important thing in relative quatification is that you need to select a sample that you would normalise others to. This one is called a calibrator and you have to choose it depending on your experiment setup, it can be control sample, it can be healthy sample, or it can be zero timepoint in time-course experiments. Selecting right calibratior is crucial for the right analysis. (the calibrator is called "control" in the Pffafl paper, but since it doesn't need to be a control sample, calibrator is better term)

You will add the values to the equation as mentioned in the paper (the real samples, no standard dilutions). Remember that calibrator must be on the same plate as all other compared samples for a single gene (if you have to many genes, you have to put calibrator on the new plate again with the rest of them). Different genes can be on different plates and it doesn't matter, because they are then normalised anyway.

If you look at your reference gene values, they should be pretty similar if you used the same amount of RNA in them (which you should). Those only slight differences between them are then used to normalise samples between reference and target gene (first normalisation), and this value is then normalised to the calibrator sample (second normalisation). The final ratio output of analysis then would be always 1 for the calibrator (or 100 % if you like) and higher, lower than 1 for the samples, that expres fold increase/decrease in the transcription of your target gene normalised to the reference and normalised to the calibrator sample (and in Pffafl case also efficiency corrected).

Dear
Trof


Thank u very much for ur help.
Realy I was confused about th term control??
is it my control sample (cDNA) or control gene (GADH) or control for my test.
Thanks alot u helped me so much

Yes, there is sometimes a confusion about that.
Since you need a gene whose expression doesn't vary between your samples, and also need a sample which will be taken as 100 % in the comparison, I like to use terms reference gene and calibrator sample (or just reference and calibrator) in this respect to make clear what are we talking about, this naming is taken fprobably from LightCycler quatification manual, I'm not really sure now.
Control on other hand can be a type of sample, which would be used as calibrator or not, because calibrator doesn't necessarily need to be a control sample. It would be better to avoid using term "control" for a reference gene at all. So mostly if anyone writes "control" it is probably a sample, but it's better to clarify that.