Hello Adam,

In my view, choosing of calibrator depends on the expression status of the target genes in bone tissue. If there are any literature supporting the same, it would be beneficial to choose.

e.g., If the target genes go undetected in the bone tissue of both normal and diseased pigs, then the calibrator which consists of pooled cDNA would be helpful, otherwise a calibrator from either one of the control or diseased should be good choice. I have performed experiments that way!! It depends on what your initial hypothesis was!!

Your analysis is valid from what you have stated.

I agree with your supervisor and with Jeru. You data will be relative to your normalizer (endogenous/housekeeping gene) and the RNA quantity from your calibrator curve.

ΔCt target = Ct target (calibrator) - Ct target(sample)

ΔCt normalizer = Ctnormalizer (calibrator) - Ctnormalizer(sample)

Fold difference = (Etarget)ΔCt target /(Enormalizer)ΔCt normalizer

Jeru and Andre, many thanks for your answers. The main reason this was brought up was because my other supervisor asked what the samples had been normalized to. But because our calibrator was a pooled sample from both healthy and diseased groups it would therefore be impossible to determine what the fold changes in our samples are relative to? 

Andre we have indeed used a housekeeping gene on every plate, but we are not determining absolute RNA quantity using a standard curve; we are using relative quantification. In this case the normalizer that I was advised to use consisted of both disease and healthy samples; therefore the fold difference in the target is relative to this pool, and not to for example say a calibrator consisting of controls of which the fold difference would have been relative to a control. Both individual disease and control samples have been ran and calibrated using this calibrator, therefore would it be feasible to present the data as being relative to each other? e.g.) disease 1 =0.8, control 1 = 0.4, disease 2 = 1.2 control 2 = 0.4, make everything relative to control 1===> disease 1=2, control 1=1, disease 2=3   control 2 = 1? 

I hope this helps at all. And thanks once more for your very helpful feedback. 

Adam I understood your problem. But it doesn't matter that your calibrator is a mixture because your curve will be made by a specific amplicon. You will have a normalized target (target/endogenous) for your test sample and for your calibrator. In the end you calculate fold difference for both samples. It's difficult to explain here. Check the pdf file I'm attaching (pages 44-47).

Yes I now understand what you are explaining Andre, many thanks for your informative feedback. The attached pdf has also proved an extremely useful read. 

 I wish you all the best of luck in your research endeavours. 

Dear Andre and Adam,

Thank you for the acknowledgement. Adam I understand your query. I just like to add on to one of your statements regarding your experiment not being standard curve. Even for a Relative quantification experiment I do a standard curve along in atleast one plate, this is done just to ensure the efficiency of the Q-PCR falls within the acceptable range. You might have done it already, I just wanted to suggest for your comment.

All  the best for your study to progress well.

Dear Andre and Jeru,

Then, I had a question, in my project, I choose age-, sex and location matched healthy controls . therefore, I have used each control sample as a calibrator against his matched disease status. would you please tell me, if it is right?

thank you 

Dear Yasin,

If you are performing the experiment in a plate, there can be only one sample as a calibrator, as per my knowledge.

From my understanding of your query, correct me if its wrong. 

You have used multiple matched controls for the corresponding diseased state sample!! So, if you are using different controls as calibrators you will have a different RQ value normalised to each?? Or have you pooled the all the controls? 

If you elaborate the experiment details, I will be in a better position to understand your query!!

Hi Yasin

Those control samples you used are healthy individuals, they are your "negative control" or your "basal expression" depending on your question. The calibrator is a sample that express all your genes of interest for calculations. For exemple if you are analyzing a gene that is only expressed in sick individuals than a sample from your healthy control won't work as a calibrator. You can use a cell line as calibrator. I even prefer because you can get high amounts of RNA and you can always use the same batch.

Dear Jeru,

yes, I have multiple matched controls for the corresponding diseased state sample. therefore, relative quantities (RQ) of each sample was determined in relation to his/her matched healthy control. Also, each sample was run with it's control in the same plate side by side.

 Dear Andre,

You are right. but, our gene of interest have been expressed in both controls and affected cases. in your opinion, in this status, can I use each control as a calibrator against his/her diseased state? if yes. please tell me what is the most suitable statistical test to analysis of difference among groups?

thank you