On 20/03/12 23:35, Geoffrey Feld wrote:

Greetings,

I have a co-crystal structure of a protein bound to an interesting polymer -- poly-D-g-glutamic acid (poly glutamic acid with D chirality and gamma linked, rather than alpha linked). In the structure, I have density corresponding to different links, from 1 to 6 (so far). I'd like to build a ligand that has "residues" much like a polypeptide polymer so that I can refine each residue individually (hopefully in COOT). Up till now, I've been using SMILES to generate the different length polymers so I have pdbs and cif files, and then (painstakingly) modelling them into COOT,

Eeeek! It is not clear to me why you are doing that. I would simply beam in D amino acids, either with Get Monomer or Search Monomer Library (if you don't know the TLC). One I have Rted the new AA into position, I'd merge the individual molecules (into
the first one typically) renumber and change chain ids... You'll have to delete the peptide hydrogens and OXTs of course. That (AFAICS) will give you a model that should Just Work with phenix.refine. I don't follow how you would have problems with GLU - or that this method would do so.

HTH,

Paul.