[BioC] FDR and filtering

[Phguardiol at aol.com]

Hi,
I d like to know if it is necessary or recommended to filter the data according to the Mean / SD ratio for replicates before running a comparison analysis using a test with a "FDR option" like SAM ?
My understanding is that if I have an important variability between replicates then the FDR would increase and be more than let say 5% so that these genes with such a variability would not be "selected" by me ?
thanks for your comments
Philippe