I want to do energy minimization before docking for accuracy. However, when I use the refinement tool and do water refinement I notice that my bond to make my peptide cyclic breaks, so it becomes linear. Is there a way to navigate this or use another tool to do the energy minimization? This happens both to head to chain cyclization and disulfide bond cyclization
You could use haddock3 which has a protocol to generate cyclic peptide conformations.
You will have to run it locally.
Check the following example: haddock3/examples/peptide-cyclisation/README.md at main · haddocking/haddock3 · GitHub