Thanks for the detailed reply, Tristan. That clarifies things! ________________________________ From: Tristan Croll <tic20@cam.ac.uk> Sent: Tuesday, May 31, 2022 12:48 PM To: chimerax-users@cgl.ucsf.edu <chimerax-users@cgl.ucsf.edu>; Anderson, Jacob <jacob_r_anderson@hms.harvard.edu> Subject: Re: Isolde | ARG Sidechain Bond Angle Outliers Hi Jacob, For a few reasons (including this one) I don't currently recommend using models straight from ISOLDE for deposition to the wwPDB. ISOLDE uses the AMBER molecular dynamics force field, which has various subtle deviations from the restraint libraries used in classical refinement and validation (the latter are based on statistical analysis of high-resolution structures; the former on replicating energies from QM simulations of model compounds). For almost all day-to-day purposes the differences are inconsequential, but it will give you various validation outliers. The other big reason is that ISOLDE doesn't currently attempt to refine atomic B-factors. The upshot is that you should still run the model through a dedicated refinement package (e.g. phenix.refine, phenix.real_space_refine, or REFMAC) before depositing. ISOLDE does try to help you with this by writing input files with settings I've found to work well - use the command "usage isolde write" for info on how to do this. Best, Tristan ________________________________ From: ChimeraX-users <chimerax-users-bounces@cgl.ucsf.edu> on behalf of Anderson, Jacob via ChimeraX-users <chimerax-users@cgl.ucsf.edu> Sent: 31 May 2022 17:40 To: chimerax-users@cgl.ucsf.edu <chimerax-users@cgl.ucsf.edu> Subject: [chimerax-users] Isolde | ARG Sidechain Bond Angle Outliers I have been using Isolde inside Chimerax 1.3 - it works fantastic! However, when I go to validate on OneDep, invariably all the ARG have outliers, specifically the NE-CZ-NH1 bond angle. Has anyone else observed this and overcome it? Perhaps something wrong with my build?