Quick navigation:        Home   |    Site Map   ||    References   |    Biography   ||    Copyright   |    Other copyright   |    Contact us   |   
Protein structure
 

Re: [ccp4bb] Rfree statistics in SHELXL
 

Basic tutorials: 		 
 

CCP4bb navigation

CCP4bb <-- 2007 <-- October 2007 <-- 08 October 2007
Previous message:
Subject: Re: Rfree statistics in SHELXL
From: "George M {- dot -} Sheldrick" gsheldr {- at -} SHELX {- dot -} UNI-AC {- dot -} GWDG {- dot -} DE
Date: 2007-10-08 		Next message:
Subject: shutdown of ccp4 website
From: Charles Ballard c {- dot -} c {- dot -} ballard {- at -} DL {- dot -} AC {- dot -} UK
Date: 2007-10-08


Subject: Re: Rfree statistics in SHELXL
From: "George M {- dot -} Sheldrick" gsheldr {- at -} SHELX {- dot -} UNI-AC {- dot -} GWDG {- dot -} DE
Date: 2007-10-08

Boaz,

If you had used the U option in SHELXPRO to update from .res to .ins the
free R should have been included in the deposition PDB file by the B
option!

George

Prof. George M. Sheldrick FRS
Dept. Structural Chemistry,
University of Goettingen,
Tammannstr. 4,
D37077 Goettingen, Germany
Tel. +49-551-39-3021 or -3068
Fax. +49-551-39-2582


On Mon, 8 Oct 2007, Boaz Shaanan wrote:

> A related but somewhat easier problem to fix, I would think, is the report of overall-R-free value in the B option in shelxpro. The R-free value is reported in the .lst and other o/p files but is not transmitted to the output pdb. Here is an example:
>
> REMARK 3 DATA USED IN REFINEMENT.
> REMARK 3 RESOLUTION RANGE HIGH (ANGSTROMS) : 1.35
> REMARK 3 RESOLUTION RANGE LOW (ANGSTROMS) : 10.00
> REMARK 3 DATA CUTOFF (SIGMA(F)) : 0.0
> REMARK 3 COMPLETENESS FOR RANGE (%) : 93.2
> REMARK 3 CROSS-VALIDATION METHOD : FREE R
> REMARK 3 FREE R VALUE TEST SET SELECTION : RANDOM
> REMARK 3
> REMARK 3 FIT TO DATA USED IN REFINEMENT (NO CUTOFF).
> REMARK 3 R VALUE (WORKING + TEST SET, NO CUTOFF) : 0.1749
> REMARK 3 R VALUE (WORKING SET, NO CUTOFF) :
> REMARK 3 FREE R VALUE (NO CUTOFF) :
> REMARK 3 FREE R VALUE TEST SET SIZE (%, NO CUTOFF) :
> REMARK 3 FREE R VALUE TEST SET COUNT (NO CUTOFF) :
> REMARK 3 TOTAL NUMBER OF REFLECTIONS (NO CUTOFF) : 44691
> REMARK 3
> REMARK 3 FIT/AGREEMENT OF MODEL FOR DATA WITH F>4SIG(F).
> REMARK 3 R VALUE (WORKING + TEST SET, F>4SIG(F)) : 0.1635
> REMARK 3 R VALUE (WORKING SET, F>4SIG(F)) :
> REMARK 3 FREE R VALUE (F>4SIG(F)) :
> REMARK 3 FREE R VALUE TEST SET SIZE (%, F>4SIG(F)) :
> REMARK 3 FREE R VALUE TEST SET COUNT (F>4SIG(F)) :
> REMARK 3 TOTAL NUMBER OF REFLECTIONS (F>4SIG(F)) : 35043
>
> Best wishes,
>
> Boaz
> ----- Original Message -----
> From: "George M. Sheldrick"
> Date: Monday, October 8, 2007 11:16
> Subject: Re: [ccp4bb] Rfree statistics in SHELXL
> To: CCP4BB@JISCMAIL.AC.UK
>
> > SHELX is much older than the free R so this was added as an
> > afterthought,
> > and unfortunately I did not think of calculating R-free for just
> > the outer
> > resolution shell. There isn't an easy way of doing this without
> > some
> > programmeing effort. I also think that it is NOT a good idea, in
> > the outer
> > resolution shell you may only about 100 reflections in your free
> > R set, so
> > the free R value will have a very large esd and so could be
> > misleading.
> > George
> >
> > Prof. George M. Sheldrick FRS
> > Dept. Structural Chemistry,
> > University of Goettingen,
> > Tammannstr. 4,
> > D37077 Goettingen, Germany
> > Tel. +49-551-39-3021 or -3068
> > Fax. +49-551-39-2582
> >
> >
> > On Mon, 8 Oct 2007, Oliv Eidam wrote:
> >
> > > Dear all,
> > >
> > > I used SHELXL for anisotropic structure refinement of high-
> > resolution data
> > > (1.3 Angstroem) and a reviewer asked me to include refinement
> > statistics for
> > > the outer most resolution shell in the crystallographic table.
> > While R-factors
> > > for the outer most resolution shell belong to the standard
> > output in REFMAC or
> > > CNS, it seems to me that SHELXL calculates R-factors only for
> > reflections of
> > > the work set. I also noticed that test reflections are missing
> > in the
> > > structure factor file (.fcf file) generated by shelxpro,
> > preventing me to use
> > > the [S] or [L] options in shelxpro to calculate R(free) statistics.
> > > Is it simply not possible to calculate the R(free) factor for
> > the outer most
> > > shell after SHELX refinement? Or is there a problem with my
> > SHELX program?
> > >
> > > Many thanks for your answers and suggestions,
> > >
> > > Oliv
> > >
> > >
> > > --
> > > Oliv Eidam
> > >
> > > Department of Biochemistry
> > > University of Zurich
> > > Winterthurerstrasse 190
> > > 8057 Zurich, Switzerland
> > >
> > > Phone: 0041-44-635-6563
> > > Email: eidamo@bioc.uzh.ch
> > > Web: http://www.bioc.uzh.ch/gruetter
> > > NCCR: http://www.structuralbiology.ethz.ch
> > >
> > >
> >
>
> Boaz Shaanan, Ph.D.
> Dept. of Life Sciences
> Ben-Gurion University of the Negev
> Beer-Sheva 84105
> Israel
> Phone: 972-8-647-2220 ; Fax: 646-1710
> Skype: boaz.shaanan‎
>

CCP4bb navigation

CCP4bb <-- 2007 <-- October 2007 <-- 08 October 2007
Previous message:
Subject: Re: Rfree statistics in SHELXL
From: "George M {- dot -} Sheldrick" gsheldr {- at -} SHELX {- dot -} UNI-AC {- dot -} GWDG {- dot -} DE
Date: 2007-10-08 		Next message:
Subject: shutdown of ccp4 website
From: Charles Ballard c {- dot -} c {- dot -} ballard {- at -} DL {- dot -} AC {- dot -} UK
Date: 2007-10-08


ProteinCrystallography.org: Copyright 2006-2007 by Quid United Ltd