I was wondering whether anybody has successfully run the DEN refinement under phenix yet. I am having some problems with my scripts stalling without any error messages. 

What sort of reference models do people use? The following issues I am not quite clear on:

- if I don't have a reference model should I call on my current pdb where the script calls for a reference model?

- is it ok to use a single domain as a reference model or is it not enough to get the whole protein refined?

- is it possible to use more than one reference model; in a case where I had structures of each domain solved separately?

- can a reference model be a different protein or does it have to have the same sequence? 

Thanks a lot!