Question: Computing fold change values for RT-PCR
gravatar for Zara
5.6 years ago by
Zara20 wrote:

Hello to everyone,

I have output from RT-PCR (Applied Biosystems - DataAssist software). The output is an excel sheet containing CT values, Delta CT, and 2 ^ (-CT) for each sample.

For each gene I have a list of samples. For each sample I have three groups (control(or reference), Adenoma and Hyperplastic). For each sample I also have the CT value of target gene and selected control. 

So each sample is categorized under one of these three groups (and values are available for target gene and selected control).

Now it's easy to compute "Delta CT" for each sample.

I just don't know how to compute "Delta Delta CT" for each group.

For example I have 9 controls, 14 Hyperplastic and 30 Adenoma. Now how "Delta Delta CT" should be computed and How a total fold change value should be computed for it?

* I know Data Assist software also does this for me, but I need to recompute it for some purpose.

Any help is highly appreciated.



fold-change rt-pcr • 25k views
ADD COMMENTlink modified 5.6 years ago by Devon Ryan98k • written 5.6 years ago by Zara20
gravatar for Devon Ryan
5.6 years ago by
Devon Ryan98k
Freiburg, Germany
Devon Ryan98k wrote:

Assuming you really do need to compute this entirely yourself, the steps are as follows:

  1. For each sample, compute the delta Ct, which is Ct_target - Ct_reference
  2. Pick a reference sample, typically one of the controls
  3. The delta delta Ct for each sample is then its value from step 1 minus the value from step 2
  4. The fold-change is 2 raised to the result of step 3.
ADD COMMENTlink written 5.6 years ago by Devon Ryan98k

Thanks a lot Devon. So it doesnt matter for which sampl which control to pick? Or it's better to average over all controls... Thank you :)

ADD REPLYlink written 5.6 years ago by Zara20

For statistics you'd average across groups (well, you'd do a T-test or ANOVA), for computing the ddCt one typically just picks a sample. This allows you to spread samples across multiple plates and just load the same reference sample in each (this isn't perfect, but often suffices).

ADD REPLYlink written 5.6 years ago by Devon Ryan98k

I have another question:

Now I have a fold-change value for each sample. Suppose I have three groups (control, treated-1, treated-2). For each group I have several samples. How can I get a single fold-change value for each group?

I saw for example R package HTqPCR that also reports a p-value for the final fold-change.
Can you help me with these two terms also?

Thanks a lot.

ADD REPLYlink written 5.6 years ago by Zara20
  1. mean or median are typically applicable to any groups of numbers
  2. A T-test or Wilcoxon test are commonly used to compare groups of values
ADD REPLYlink written 5.6 years ago by Devon Ryan98k

Thank you Devon ^_^

ADD REPLYlink written 5.6 years ago by Zara20
Please log in to add an answer.


Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 2.3.0
Traffic: 1047 users visited in the last hour