Dear all, I am refining the structure of a cyclic peptide with having residues that are not in the ligand list of phenix. I would to perform unrestrained refinement by considering the very good resolution of the data. However, also in this case, phenix requires the restrain file of the ligand. I got cif file in two different way: 1) PRODRG server 2) by running phenix.elbow both produce the restrain file with no error message. However, by using such file in the refining, I get this error: Sorry: Fatal problems interpreting model file: Number of atoms with unknown nonbonded energy type symbols: 14 In the cif file, there are not atoms having 14 as energy type symbols. Could you suggest to what this message is referred to? How can I modify the cif to overcome the problem? Thanks in advance, Danilo -- Dr. Belviso Benny Danilo, PhD Istituto di Cristallografia (IC) Consiglio Nazionale delle Ricerche (CNR) via Amendola 122/o, 70126 Bari - Italy Tel ++39 080 5929273 Fax ++39 080 5929170 mail: [email protected] web: users.ba.cnr.it/ic/crisdb00