I have tried to do build single point mutant models (n=100) along with separate optimization step on the neighbor residues around the mutated residue.I can able to get 98 models without any issue out of 100 models but in some models, the bond lengths of longer residues(ARG,LYS,TYR) became extreme than original value.
For example: IN ARG residue, the distance between (CZ & NH1) or (CZ & NH2) was extreme (3.4Å) than the original value (1.5Å ) which makes the bonds disappear.
I really don't understand that what makes these issue and I use the default MD&SA parameters which used in mutate_model.py with Lenard-Jones potential. *I don't get this issue if I change the restraint from Lenard-Jones into soft-sphere potential.
So, kindly give me some suggestion about the parameters that I can tune to resolve this issue.
Thanking you in advance
With Regards Mahesh V