I am having an issue with secondary structure display. I am real-space refining a set of 120 chains in phenix. Of these, only 10 are unique, the rest is symmetry-related. Thus, I define secondary structure elements for the 10 unique chains only. The output pdb file header contains the secondary structure elements as imposed during refinement. Displaying this pdb file in ChimeraX shows the right secondary structure. The mmcif file instead, whether I use the phenix output or generate it from the pdb myself, displays different, non-sensical structures. Secondary structure is shown in loop regions, and some actual strands and helices not recognized at all. This artifact is only shown for the 10 subunits with secondary structure definitions, the other subunits show no secondary structure at all, thus something is read in, but not correctly, apparently. Any idea what the problem could be? Could the size of the complex be an issue?

The attachments show the lines from the two file types concerning secondary structure definitions.

Kind regards,

Stephan