Take the 2-minute tour ×
Biology Stack Exchange is a question and answer site for biology researchers, academics, and students. It's 100% free, no registration required.

I want to evaluate the level of gene expression by real-time PCR. I have five controls that are "clinically" the same. I calculated the "fold change" of the target gene regarding each control.

What if the result of "fold change" for my target gene is different for each control? I mean it is up regulated according to three of them and down-regulated according to the rest two.

Is the intervention biased? Or should I consider the average?

share|improve this question
    
I have no experience with RT-PCR in particular, but I'd say that the variations in your controls add to your error. And if they vary too much, you'll have a hard time to measure a significant difference in gene expression. The gene you selected might be influenced by a lot of other processes, leading to those variations. –  Mad Scientist Apr 5 '12 at 20:21

2 Answers 2

You also probobably need to check if your samples haven't been contaminated with PCR reaction inhibitors, which is very common if you first extract your mRNA, digest remaining DNA and then run a PCR with a less then 10-fold dilution. You need at least a 200 times dilution to get rid of all the artifacts.

Once you've diluted your samples, place a repeats of successive dilutions of your target and baseline mRNA (200xdilution, 1000x dilution, 5000x dilution). Normally you will see a nice regular spacing between curves on your rtPCR plot. If you don't, something went wrong with the reaction.

More generally rtPCR is a pretty tricky experience, and if you want it to be 100% reliable, you've got a certain number of controls to perform. You can find a good and comprehensive guideline over here.

share|improve this answer

Please provide more information: fold-change relative to what?

If you did what I think you did (single control gene that you calculated fold change to of your gene of interest) I'd say this is the wrong approach. What you need is a set of genes which have similar expression levels across all your samples (controls and cases) to be able to compare your gene of interest to some common baseline. Selection of such genes should be the first step in the project, and it might be a good idea to use one of the established approaches - I recommend Jo Vandensompele's GeNorm (link to the paper) method. It goes like this: from a panel of potential control genes you select two or three that are most stably expressed across your experimental conditions, and then use these two-three control genes along with your gene of interest in a qRT-PCR. You then normalize the signal from your gene to a mean of the control genes.

It has been repeatedly shown that using a single control gene, even a so-called 'housekeeping gene' is misleading, because those genes do change their expression levels in some conditions. Also, using several control genes protects you against potential problems due to unequal amplification efficiency in different samples/groups. However, if you have multiple gene of interest, they all must have comparable amplification efficiency.

share|improve this answer

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.