Hi Pavel, I am refining a structure to 1A resolution (which I had asked you about before). It is nearly complete. However, one peptide bond is very short (1.2A vs ~1.33A). From the density, it appears that the two residues flanking the peptide bond have (at least) two conformations. I have tried to model both as a single and double conformers. In both cases, PHENIX seems to force the single or both conformers into the average position, which of course leads to the short apparent peptide bond. Can you suggest any way to fix this problem? If you prefer, I can send you the models and the data file to have a look. Thanks a lot. Jeffrey