Hi All, I have Chip-Seq data for histone marks (H3K27me3, H3K27ac, H3K4me1 and H3K4me3) for 2 different cancer subtypes which differ in their aggressiveness. I am interested to see if there is any association between the tumor aggressiveness and H3K27me3 loss.

I have used Diffbind to call differential binding sites between different subtypes for each antibody. I read it somewhere that the H3K27ac shares its binding location with H3K27me3 and often seen interacting with H3K4me3 in bivalent domain (please correct me if I am wrong).

With the given background, I am planning to compare the coordinates that has less binding for H3K27me3 and are compensated with any of the activation marks like H3K27ac, H3K4me1 or H3K4me3 to specifically call sites with H3K27me3 loss.

I am not sure whether it will be a correct approach to compare the exact coordinates to find out the sites with loss of H3K27me3 and or it should be addressed differently?

Any suggestions will be highly appreciated.

Thanks