Hi Jeff, here is the solution: 1) Rename your unknown amino acid from UNK to something like ZZZ (since UNK has special meaning); 2) Run elbow.builder to generate a cif file (stereochemistry definitions) for this unknown item: % elbow.builder model.pdb --do-all 3) This will create a cif file for ZZZ (the file you need named like elbow.model_pdb.all.001.cif, you can remove the other elbow output); 4) Run phenix.refine again specifying this cif file from the command line: % phenix.refine model.pdb data.mtz elbow.enk_zzz_pdb.all.001.cif If you have multiple unknown items, call them differently like ZZ1,ZZ2, ... etc, so elbow.builder do not consider them as one single ligand. Also, make sure you specify element type for each atom in unknown entry in your pdb file (column 78), so it looks like : ATOM 32 CA ZZZ A 5 3.055 1.059 3.693 1.00 0.87 C ATOM 33 CB ZZZ A 5 3.965 0.435 2.634 1.00 1.13 C ATOM 34 CG ZZZ A 5 3.531 0.603 1.177 1.00 1.16 C ATOM 35 CD1 ZZZ A 5 3.411 2.076 0.818 1.00 1.88 C ATOM 36 CD2 ZZZ A 5 4.502 -0.103 0.245 1.00 1.67 C ATOM 37 C ZZZ A 5 1.634 0.527 3.541 1.00 0.87 C ATOM 38 N ZZZ A 5 3.576 0.800 5.030 1.00 0.92 N ATOM 39 O ZZZ A 5 1.246 -0.440 4.196 1.00 1.23 O ATOM 40 OXT ZZZ A 5 0.073 1.000 3.735 1.00 1.11 O Please let me know if you have any more questions. Pavel. Jeff Habel wrote:
What's the best way of dealing with unknown atoms (UNK) in phenix.refine?
I have a long disordered loop that runs basically the length of the protein on the back side and there is a single "amino acid" in the middle that I'm trying to model as UNK. I can run a round of refmac and auto generate a lib file with my ligand and the unk atoms but phenix.refine balks at that with respect to the Van der Waals radii.
Thanks for the help,
-Jeff
------------------------------------------------------------------------
_______________________________________________ phenixbb mailing list [email protected] http://www.phenix-online.org/mailman/listinfo/phenixbb