I'm new to microarray and I'm trying to analyze several single channel agilent microarray using Limma package.
I have a question concerning the calculation of FDR q value using eBayes/topTable.
If I understood well, q value is an adjustment of p value linked to gene ranking. So, should the q value be calculated on an entire dataset or should it be calculated after filtering (removing non-coding RNA for instance)? As the number of variables will be different, the ranking should also be, will this impact the q value?