[phenixbb] MTZ file

Pavel Afonine PAfonine at lbl.gov
Tue Aug 26 12:37:05 PDT 2008


Hi Yu Seon Chung,

- I'm not sure I understand why you don't want to use your original file 
with reflections...

- If you start the refinement with non-MTZ file (for example, CNS or 
SHELX) or the input data was scattered across multiple files (Fobs in 
one file, HL coefficients in another file, free-R falgs are generated or 
in third file, etc...), then phenix.refine consolidates all input data 
and writes it out into a file ending with "_data.mtz". This is the file 
that you should use in future runs. The data in "_data.mtz" is not 
processed, that is no sigma or resolution cutoffs are applied to it 
(even though you used these cutoffs in refinement).

- phenix.refine has an option to output calculated structure factors, 
such as Fmodel, Fcalc_atoms, Fbulk, Fmask, and some other scales, Fobs, 
free-R flags, etc... The data in this file may be truncated by 
resolution, sigma, etc. This file is not output by default 
(export_final_f_model = mtz cns); you need to place "*" in front of cns 
or mtz to request its output. It is not assumed that you will use this 
file in next refinement runs. And I presume this is the file you are 
looking at.

Pavel.

 
On 8/26/2008 11:02 AM, YS. Chung wrote:
> Hi all, 
>
> I have a simple question about the output mtz file from phenix.refine.
> When I look at the mtz file, it says 
>
> ***********************************************************
>                     
>  DO NOT USE THIS FILE AS INPUT FOR REFINEMENT!
>                                   
>  Resolution and sigma cutoffs may have been applied to FOBS.
>                     
>  ***********************************************************       
>
> If I have not applied either resolution or sigma cutoffs to FOBS, does
> that mean that this mtz file is ok to be used for the next round of
> refinement? 
>
> Thank you,
> Yu Seon Chung
> _______________________________________________
> phenixbb mailing list
> phenixbb at phenix-online.org
> http://www.phenix-online.org/mailman/listinfo/phenixbb
>   



More information about the phenixbb mailing list