I have noticed the same thing for such high resolution maps… so I lately I have been depositing ISOLDE refined models. I don’t think there is a strict reason to deposit models refined with one method or another. As to the reason, it could be that Phenix uses stronger restraints for backbone dihedrals (i.e. may have lower Ramachandran outliers) and rotamers (less unfavored rotamers). But this may make side chains not fit the density as well. On the other hand, ISOLDE models seem to fit the density better and have lower clash scores. If you want to stick with Phenix there are probably some parameters you can tweak to make these restraints not as strong perhaps? Greg
On Dec 4, 2025, at 3:02 PM, man6ny--- via ChimeraX-users <chimerax-users@cgl.ucsf.edu> wrote:
Dear all,
Maybe an ignorant or previously-asked question... I believe most people refine cryo-EM models using e.g. Phenix real-space-refine before depositing them to the PDB. Is there a strict reason (practical or philosophical) why cryo-EM models refined in ISOLDE cannot be directly deposited without subsequent refinement in Phenix or Refmac? I have a high-resolution map (2.2 A) but the latter programs seem to actually make my model worse (e.g. Phenix tends to give me rotamers that are not supported by the experimental data). I would prefer to use the model refined in ISOLDE but perhaps I am missing something?
Many thanks for your insight!
Louis _______________________________________________ ChimeraX-users mailing list -- chimerax-users@cgl.ucsf.edu To unsubscribe send an email to chimerax-users-leave@cgl.ucsf.edu Archives: https://mail.cgl.ucsf.edu/mailman/archives/list/chimerax-users@cgl.ucsf.edu/