I've run DGE in edgeR between two sample groups with 3 replicates each and tabulated a list of DEGs.

I also have a positive and negative control gene in each sample that I am aiming to use as a baseline for further investigation. I want to pare down to genes which are consistently greater in expression than the baseline controls.

Is it meaningful to take the difference in logFC between DEGs and the controls of interest as a metric for this?--or is that simply statistically unsound?

If this doesn't make sense, is there a way to incorporate the controls in the modelling to account for this goal?