[phenixbb] Increase in R/R(free) after phenix.real_space_refine

Pavel Afonine pafonine at lbl.gov
Fri Sep 19 13:32:17 PDT 2014

Hi George,

the target function implemented for real-space refinement is such that 
it moves atoms towards closest map peaks (the target is the negative sum 
of map values calculated at atomic centers). Obviously, this target is 
not very similar to R-factor function or reciprocal space targets that 
are more or less quadratic near the minimum. All in all this means that 
real-space refinement does not guarantee R-factor decrease and a 
following round of reciprocal space refinement is necessary if you plan 
to look at R-factors. The reason why this target is use is because it is 
fastest to calculate.


On 9/19/14 1:22 PM, Nathaniel Echols wrote:
> Global real-space refinement simply isn't very useful at high 
> resolution, and you shouldn't need a large radius of convergence when 
> the R-free is already 13%.  I'm a big fan of automating everything 
> possible, but this is one case where fixing the sidechains manually in 
> Coot (which should do an excellent job) is by far the easiest option. 
>  Once you fix the sidechains the C-beta outliers will probably disappear.
> PS. A common reason for C-beta deviations at high resolution is 
> residues with alternate conformations being split at C-beta rather 
> than splitting the entire residue, which allows the backbone to move 
> to accommodate the optimal placement of the sidechain (rather than 
> straining the geometry to get the sidechain atoms to fit the density). 
>  Always split the entire residue - the backbone is almost never rigid 
> in disordered regions.
> On Fri, Sep 19, 2014 at 1:11 PM, George Devaniranjan 
> <devaniranjan at gmail.com <mailto:devaniranjan at gmail.com>> wrote:
>     Hi,
>     I refined (phenix.refine)  a 1.0Å structure but it had a few CB
>     outliers and some rotamer outliers.
>     I decided to try real_space_refine, to check if they could be
>     fixed (given that it has a higher radius of convergence).
>     (using run=local_grid_search+minimization_global in real_space_refine)
>     But, the R/R(free) increased from 11.39/13.24( for the refine
>     model) to 16.51/17.55 (real_space_refine model).
>     It seemed a big increase, when I looked at the map against the
>     structure the protein itself seemed fine.
>     The rotamers outliers had decreased, the CB outliers were fixed..etc
>     The solvent fit to the density had worsened in COOT.
>     What am I doing wrong? Any suggestions, advice would be much
>     appreciated.
>     Thank you,
>     George
>     _______________________________________________
>     phenixbb mailing list
>     phenixbb at phenix-online.org <mailto:phenixbb at phenix-online.org>
>     http://phenix-online.org/mailman/listinfo/phenixbb
> _______________________________________________
> phenixbb mailing list
> phenixbb at phenix-online.org
> http://phenix-online.org/mailman/listinfo/phenixbb

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://phenix-online.org/pipermail/phenixbb/attachments/20140919/6ce7df86/attachment.htm>

More information about the phenixbb mailing list