Differential network analysis

Hi All,

I want to perform differential correlation analysis for drug treatment. Is it to alright get the normalized microarray values then directly perform the differential correlation between the control and drug treatment? Or is it more “correct” to get the fold change values (relative to control/water treatment) first then perform the differential correlation analysis? If i do the latter i can compare the drug treatment i want vs. random treatments as controls

Many thanks for your input!