I have a protein structure with 394 amino acids.
I tried to add some missing residues from aa 395-400, obviously at the terminal end. It was successful.

Q1:

After loop refinement (non-missing residues did not move), the global DOPE score was improved, however PROCHECK analysis showed a minor decrease ( <1% only) in the overall quality of the model. I also noticed that the added residues were 'pointing' outwards in all the built models. Before refinement, they were close to other residues.

In this case, does it mean that loop refinement is not necessary?
Or just ignore PROCHECK analysis? Since I'm not building a new homology model, but insertion of few residues only.

Q2:

In homology modeling, does DOPE score always correlate with the quality of stereochemistry? 

Hope to get some feedback and learn from you all.
Thank you.

Regards,

Adam Lim

Institute of Biological Sciences,

Faculty of Science,

University of Malaya,

Malaysia.