[phenixbb] on graphical interface of Phenix 1.10.1-2155-000

Pavel Afonine pafonine at lbl.gov
Wed Jan 25 07:45:17 PST 2017


Hi John,

> For the "refinement settings" in the phenix.refine graphical 
> interface, will you please let me know when we need to change the 
> parameters in "Refinement target weight..."? And is any rule on how to 
> modify the  parameters in "Refinement target weight..."?

normally, you only need to change anything in that department if you 
think it is necessary. If you don't know or doing sort of "SA vs NCS vs 
TLS vs twinning" experiments then it's better leave that alone.

Enabling weight optimization makes sense if:
- you are doing the final round of refinement (say before PDB deposition);
- have enough CPU power or happy to wait longer for hopefully better result;
- have a strong reason to believe that automatically determined weights 
are not optimal (for example, Rfree-Rwork gap is too large, or have many 
geometry outliers, like Ramachandran plot outliers, etc).

See also:

Improved target weight optimization in phenix.refine
http://phenix-online.org/newsletter/CCN_2011_07.pdf

All the best,
Pavel

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://phenix-online.org/pipermail/phenixbb/attachments/20170125/fcaf188b/attachment.htm>


More information about the phenixbb mailing list