Dear all, 

i would appreciate please a piece of advice please : what normalization method would you recommend in the following situation :

someone did ChIP-seq experiments on protein X :

1) in cells that have the INTACT protein (control cells) and

2) in the cells where the protein X was KNOCKED-OUT (we still a residual level of protein X and a residual level of X binding to chromatin)

thanks, 

bogdan

Because you are expecting a global change in the signal for X, I would recommend using the background normalization method from csaw based on read counts in very large bins (see section 4.2.2 of the csaw User's Guide). Keep in mind that with this method, you are assuming that the pulldown efficiency of your antibody is consistent across multiple samples, since variation in pulldown efficiency is indistinguishable from a genuine global change in protein binding. If you have technical replicates, you can assess how consistent your antibody is.