Hi Clemens, thank you for looking into this! To me, a "structure" that has a few sequence shifts does not appear useful at all - rather, it taints the PDB. This one should be blacklisted, like 6MO[0,1,2]. It's painful to learn that despite all the validation that is performed before and at deposition, such entries slip into the PDB !
Can PDB_REDO or validation tools identify or even correct such errors? I guess they cannot, currently. Probably it's difficult to diagnose the real problem underlying Ramachandran outliers and bad density. Could the RCSB, as part of validation, run a pipeline (like ARP/wARP, phenix.autobuild, Buccaneer/refmac) that uses the submitted model only to generate starting phases, and iteratively builds a new model that is then compared to the submitted model? I guess that could be done, at least as long as the resolution is high enough - how high is high enough?. In my experience, in cases where one can correct errors based on the deposited structure factors, also the raw data are often suboptimally processed - the existence of expert-visible errors in the end result suggests that many upstream decisions can and should be improved as well. The only way to improve the situation would be deposition of raw data (Zenodo, SBGrid, proteindiffraction.org, ...), with a link in the PDB. best, Kay On Thu, 1 Aug 2019 10:03:49 +0100, Clemens Vonrhein <vonrh...@globalphasing.com> wrote: >Hi Ivan, > >On Wed, Jul 31, 2019 at 05:32:24PM -0400, Ivan Shabalin wrote: >> And Rfree of 36% seems really high. > >If you look at the maps (e.g. after some re-refinement with your >favourite refinement package) it seems as if there are a few sequence >shifts (around and after A78), some poor density and additional >unmodelled density ... which all add to those high R-values I guess. > >But given the poor data quality and no raw images (to check and maybe >improve upon that), I didn't feel the urge to delve into that any >further ;-) > >Cheers > >Clemens > >-- > >*-------------------------------------------------------------- >* Clemens Vonrhein, Ph.D. vonrhein AT GlobalPhasing DOT com >* Global Phasing Ltd., Sheraton House, Castle Park >* Cambridge CB3 0AX, UK www.globalphasing.com >*-------------------------------------------------------------- > >######################################################################## > >To unsubscribe from the CCP4BB list, click the following link: >https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB&A=1 ######################################################################## To unsubscribe from the CCP4BB list, click the following link: https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB&A=1
