[phenixbb] PHENIX-Coot problem

Marcelo Carlos Sousa Marcelo.Sousa at colorado.edu
Wed May 7 14:05:55 PDT 2014

I do not quite understand what is going on but I fixed it:

It turned out that if you point phenix (in preferences) to the wrapper script inside the coot application package:


and start the application by clicking on the Phenix icon and then opening refinement results in coot, then the coot dictionaries do not load and there is a problem

However, if you start phenix form the terminal then everything is fine.

Finally, if you point phenix to a different file in the coot application package:


then things work fine no matter where you start phenix from…

I am on MacOS 10.8.5 and this happens with phenix 1.8.4-1496 as well as 1.9-1690…


On May 7, 2014, at 10:57 AM, Nathaniel Echols <nechols at lbl.gov<mailto:nechols at lbl.gov>> wrote:

On Wed, May 7, 2014 at 8:30 AM, Marcelo Carlos Sousa <Marcelo.Sousa at colorado.edu<mailto:Marcelo.Sousa at colorado.edu>> wrote:
I recently updated to CCP4 6.4 including Coot.
I updated the path in phenix to point to the coot wrapper startup script.
Phenix then correctly starts Coot and both programs are "connected". However regularizations or real space refinement fail with the "Failed to find restraints for: …" message as if the dictionaries are not loading.
This does not happen when starting Coot from the terminal or form CCP4i…
Any ideas on how to solve this problem?

Not exactly, but I have an idea about how to debug it.  Coot's terminal output is relatively verbose and should mention something about the monomer library when it is initialized (and perhaps when you try regularization).  So, could you please launch it and run a quick regularization, paste the terminal output to a text file, and send me the file?  You need to do this twice: once with Coot launched from the terminal, then launch Phenix from the terminal and start Coot from Phenix.  Hopefully there will be something missing in the output from the second try.

I will see if I can reproduce the problem here - however I do not believe we have observed anything like this in the past.  For what it's worth, Phenix does not override any environment variables or restraints pertaining to Coot - it will still use the CCP4 monomer library.  (It will load custom CIF files if you open it from phenix.refine, but I doubt this is the source of the problem.)


PS. Which OS/Phenix build?

Marcelo C. Sousa, Ph.D.
Associate Professor
Department of Chemistry and Biochemistry
University of Colorado at Boulder
596 UCB
Boulder, CO 80309
Phone: (303) 735 4341
Fax: (303) 492 8425

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://phenix-online.org/pipermail/phenixbb/attachments/20140507/797eadd9/attachment.htm>

More information about the phenixbb mailing list