[buster-discuss] free R flag questions

Claus Flensburg claus at globalphasing.com
Fri Sep 24 11:08:31 CEST 2010


Hi Wolfram,

On Wed, Sep 22, 2010 at 08:40:14PM -0600, wtempel wrote:
> Hello.
> Here are some excerpts from stdout of an initial autobuster "refine" run.
> <excpt>
>  ################################ Basic checks
> ################################
> 
> 
>  ESC[1mNOTEESC[m : running with the following command line arguments:
>         "-p ../201009222206.pdb -m ../output_PHENIX0phs2.mtz
>         -autoncs ColumnName_FreeR_flag=Free-Phenix -R 30
>         2.58"
> 
> </excpt>
> 
> During the first autobuster run, the ccp4 program unique seems to be
> invoked. And I get this warning:
> 
> <excpt>
>  ESC[1mWARNINGESC[m : had to create a fresh test-set column - usually
>            the MTZ file should already contain this!
> 
>  F/SIGF pairs : FP,SIGFP
> 
>  ##############################################################################
> 
>  # refining against FP,SIGFP with: Free-Phenix
>  ##############################################################################
> </excpt>
> 
> Can I ignore this warning regarding the "fresh test-set column"? 

yes, it should be harmless. It can be avoided by doing:

% refine ... ColumnName_FreeR_flag_allowed=" I[ ]* Free-Phenix" ...

> Will that
> new test-set column be used in any way?

no, it shouldn't.

> And related to this, is there some command line switch that allows me to
> specify a non-0.00 value that flags  "free" reflections?

well, yes, the following should do that:

% cat >> .autoBUSTER<<EOF
BusterExtraKwds="FREFLG=2"
EOF
% refine ...

to use a value of 3 for the test set reflexions.


Regards,

ClAuS + BDG


More information about the buster-discuss mailing list