Hi Ed, try using ncs_search.exclude_selection=None or ncs_search.exclude_selection="not all" or ncs_search.exclude_selection="not ((chain A and resseq 1) or (chain B and resseq 1))" Note, by default it is: "not ((chain A and resseq 1) or (chain B and resseq 1))" this is why it excludes your ligands. Pavel On 7/25/17 20:23, Edward A. Berry wrote:

Is it possible to restrain cartesian NCS for ligands in phenix.refine? I included a ligand in an paired ncs-group and it seems to be omitted from the listing of those groups in the log file. When I made an ncs_group including only ligands, it halted complaining of empty selection. Ed _______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb Unsubscribe: [email protected]

Is it possible to include a ligand (prosthetic group actually) in a rigid_body group together with protein chain? I have an iron-sulfur cluster (FES) which has the same chain letter as the protein subunit to which it is attached, and the rigid_body is defined by chain and sequence including the FES res number, but the cluster gets left behind when RB refinement yanks the chain into position. The cluster atoms are labeled ATOM not HETATM, in case that matters. In difficult cases including the cluster could be critical to success, since it will define the highest density peak. Ed