[phenixbb] help with geometry target weight

Pavel Afonine PAfonine at lbl.gov
Thu Mar 18 14:10:00 PDT 2010


Hi Breann,

> I am currently refining a structure at 2.4A resolution. I would like to increase the geometry target weight during refinement, however when I decrease wxc_scale (to 0.25 or 0.1) there is no change in the bond deviations or R factors compared to refinement with the default wxc_scale. I noticed in the refinement logfile under XYZ refinement wc=0. This does not change when I change the wc value in the command line (wc=2.0 for example) as suggested in a previous email. I also see no change when I use the optimize_wxc=true command. Examples of the command line input and output log files are below. I am currently using version 1.6-289 and would appreciate any help in resolving this issue as it seems the geometry is not being properly weighted during refinement.
>   

what you describe is extremely weird. The restraints weight scale, wc, 
is set to 1.0 by default and is never adjusted automatically unless you 
explicitly specify a value from the command line or a parameter file. 
So, I have no idea why you end up having wx=0.

The only way I can find this out if you send me the model, data, cif 
files (if any), parameter files (if any) and the exact command you run, 
so I can reproduce the behavior you see and tell what happens.

Pavel.




More information about the phenixbb mailing list