Hi Oliver, indeed, in recent versions phenix.refine tries to keep existing waters as intact as possible, but can remove them if they do not satisfy screening criteria. You can ask the program to keep existing waters unconditionally by using "ordered_solvent.keep_existing=true" (also can be set in the GUI). Pavel On 2/23/24 08:53, Oliver H. Weiergräber wrote:

Hello,

after switching from version 1.20 to 1.21, I noticed that the updated list of waters generated by the ordered_solvent option in phenix.refine is no longer numbered sequentially; it seems a final re-numbering step is now skipped. Is this intentional? Of course it can be advantageous to keep waters from changing their numbers, e.g. if they are part of occupancy groups. On a related note, is there a mechanism to protect certain "important" waters from being deleted while all the rest is updated as usual? One obvious option would be a dedicated chain ID, but afaik phenix.refine accepts a chain ID for output but not for input waters ...

Best regards, Oliver

==================================================   Prof. Oliver H. Weiergräber   Institute of Biological Information Processing   IBI-7: Structural Biochemistry   Forschungszentrum Jülich   52425 Jülich, Germany   Tel.: +49 2461 61-2028   Fax: +49 2461 61-9540 ==================================================

_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb Unsubscribe: [email protected]

Hi George, this is fixed in nightly builds. For example, this should work in dev-5430: https://phenix-online.org/download/nightly_builds.cgi Let us know if you are still having any problems! Pavel On 9/11/24 11:25, georgep wrote:

Hello,

I am trying to do a twin refinement and everything goes very well until the end when the job ends with an error about 2mFo-Fmodel maps not being allowed.  I have the map flags set to 2mFo-DFc but it does not seem to matter.

There is also an issue when trying to the GUI for a twin refinement, the twin_law entry seems to be ignored.  It is recognized when put the model tree in a Phil file, so that’s OK.

Thanks for any help.

George

...

On Feb 23, 2024, at 10:53 AM, Oliver H. Weiergräber wrote:

Hello,

after switching from version 1.20 to 1.21, I noticed that the updated list of waters generated by the ordered_solvent option in phenix.refine is no longer numbered sequentially; it seems a final re-numbering step is now skipped. Is this intentional? Of course it can be advantageous to keep waters from changing their numbers, e.g. if they are part of occupancy groups. On a related note, is there a mechanism to protect certain "important" waters from being deleted while all the rest is updated as usual? One obvious option would be a dedicated chain ID, but afaik phenix.refine accepts a chain ID for output but not for input waters ...

Best regards, Oliver

==================================================  Prof. Oliver H. Weiergräber  Institute of Biological Information Processing  IBI-7: Structural Biochemistry  Forschungszentrum Jülich  52425 Jülich, Germany  Tel.: +49 2461 61-2028  Fax: +49 2461 61-9540 ================================================== _______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb Unsubscribe: [email protected]

George N. Phillips, Jr. Associate Dean for Research, Wiess School of Natural Sciences Ralph and Dorothy Looney Professor BioSciences at Rice Rice University, MS140 6100 Main Street Houston, Texas  77005-1892 Office: +1 713.348.6951 www.phillipslab.org

_______________________________________________ phenixbb mailing list [email protected] To unsubscribe send an email [email protected] Unsubscribe: phenixbb-leave@%(host_name)s