Hello everybody

I faced a problem in analysis of the data set "GSE27274" by limma package. The parameter "Coef" which shows "logFC" calculated by limma  is different from the one I gained manually. I calculated log2 of treatment expression values average divided to control expression values average. I have not experienced this problem in analysis of other datasets. Is there any problem with my calculation?

Thank you

Yes, there's a problem with your calculation. Assuming you are just comparing two groups, you need to log2 first, then average, not the other way around.