fold change in gene expression

Question

[rob.costa1234]

I have a very basic question (apology, but get confused) In calculation of fold change in gene expression data I am doing Treated - control samples and then use the formula (IF(R2>1,R2,-1/R2)) (assuming R has the results. So my questions are : 1. Do we do Treated - control or other way around. 2. Is it OK to use (IF(R2>1,R2,-1/R2) above as it will take care of down regulation.

Thanks Rob

Answer 1

[natasha.sernova]

See this post:

How to calculate "fold changes" in gene expression?

Or this one, if you have several groups:

Fold change calculation

Answer 2

[WouterDeCoster]

I'm not sure you should be calculating these fold changes manually in excel. For RNA-seq there are far more sophisticated approaches, modelling the (technical) variation between replicates to get accurate estimations of fold change with accompanying p-values to indicate the significance of the difference. What are you trying to achieve? A valuable step by step workflow can be found here: http://www.bioconductor.org/help/workflows/rnaseqGene/

[rob.costa1234]

This is real time PCR data Theoretically should we do Treatment - control or only Treatment / control is good

[WouterDeCoster]

In the case of real time PCR data, the links provided by Natasha may prove to me more informative. Do you mean to subtract the signal for control from the signal from Treatment? I'm sure for qPCR there is also specific software available.

Please use tags accordingly to the content of your question, you listed RNA-seq but this isn't RNA-seq.