To calculate relative gene expression, I normally use the 2-ddCt method. In this, I have always subtracted housekeeper-target and then 2^(difference) and normalized against control. However, I never thought what could happen the other way.
Would they get different results? Also I noticed in the original paper by Livak and Schmittgen, 2001, they have subtracted target-housekeeper. I am confused now. Please guide me.
I have attached here a file showing both ways of calculating between target gene and housekeeper for a 4 group comparison. I am confused because both give different results. Please tell me which one to choose and why ?
As is the case most of the time, there is no simple answer. However, we have used the ddCt method for real time PCR in my lab for some time (see Jensen et al http://www.ncbi.nlm.nih.gov/pubmed/2362675 for details).
In short, you need to pick not only a control gene for this method, but it must be the right control gene - Tata Binding Protein works well, but sometimes you may need to use GAPDH or some other gene that DOES NOT CHANGE expression in your cell or tissue type. This may take some trial and error.
Once you find a gene to use for normalization, you will use gene expression for this gene to normalize across all of your genes in that qRT-PCR run. That is why they are all compared to this one gene. That normalized value can now be used to look at gene expression changes across samples for a given gene. Your are correct that it gets confusing which direction gene expression is changing! In fact, during the calculations you will square the adjusted value (thus making negative values positive). You must assign the direction of fold change (+ or -) based on the raw values - for example a gene that appears sooner in the experimental vs control (say cycle 20) vs one that appears later (say cycle 28) would be gene expression that went up ( cycle 20) vs down (cycle 28) if the gene expression indicated that the gene appeared at cycle 25 in the control sample. Does that make sense (and yes, I left a lot of detail out here, but have an excel sheet I can share if you like with all of the equations in it already).
Good luck!
Hi Lawrence,
Thanks for your answer and sorry for the delay in replying. I usually use up to 3 housekeeper genes like gapdh, rpl13a, polr2b or others. The crossing points of my target gene are higher (around 23 or 24 here) compared to that of housekeeper (gapdh in this case around 17).
Please send me the excel sheet and also the full text of the article (link provided by you but I do not have access to it). My question is not yet resolved, may be excel sheet helps me this time.
hello rodhan,
Please check this out It has the explaination for your question regarding substraction.See the section describing "ΔReference Gene vs ΔTarget Gene" in the following link...
http://blog.mcbryan.co.uk/2013/06/qpcr-normalisation.html
Sorry, Roshan - that was the wrong link! Here is the paper at PLOS - open access:
http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0061952
- Badges
- Science method