[phenixbb] Simulated annealing composite omit map

Hello All, I am trying to make the simulated annealing composite omit map for a protein ligand complex at 4.37A resolution. I think, Simulated annealing is harsh and rendering the protein distorting a lot as is evident from the helix turning into the some random structure. In order to avoid this protein distortion in protein backbone, do I need to fix some weight or specific parameter or provide reference structure and put harsh weight ? Any advice or suggestions would be much appreciated. Appu