Hi Yasser, could you please send me model and data and indicate what loop you are talking about, then I will have a look and comment. Pavel On 2/2/17 06:35, Yasser Almeida Hernández wrote:

I am refining a structure at 1.9A, which contains a 7 residues C-terminal region. The resolution of the map in that region is low but is clear enough to define some side chains and the backbone. I build this region manually in Coot. This region is disconnected from the core of the protein, since the connecting loop cannot be seen in the density map. After running phenix.refine (GUI) and playing with the options several times, the final model of this region is always out of the map, even when the R values are improved. How can I refine the structure, keeping that part of the protein in its place, fitted to the map?