Hi Joao,
The problem may come from your reflection file. For instance, when
you convert a .sca (obtained using Scalepack) containing intensities into a
.hkl using to_cns, you get Is and Fs with Fobs=0 for reflections with
negative Iobs. I guess these reflections are rejected even with a sigma
cutoff of zero. Try to apply rejection criteria to intensities as in the
following example:
{* reciprocal space array containing observed intensities: optional *}
{* required for the "mli" target *}
{===>} obs_i="iobs";
{* reciprocal space array containing sigma values for intensities: optional *}
{* required for the "mli" target *}
{===>} obs_sigi="sigi";
...
{* apply rejection criteria to amplitudes or intensities *}
{+ choice: "amplitude" "intensity" +}
{===>} obs_type="intensity";
{* Observed data cutoff criteria: applied to amplitudes or intensities *}
{* reflections with magnitude(Obs)/sigma < cutoff are rejected. *}
{===>} sigma_cut=-3.0;
Hope this helps,
Jean-Paul
___________________________________________________________
Dr Jean-Paul RENAUD
CNRS UPR9004 Biologie structurale
Institut de Genetique et Biologie Moleculaire et Cellulaire
1 rue Laurent Fries
B.P. 163
F-67404 Illkirch Cedex
FRANCE
Tel: +33 (0)3 88 65 33 48
Fax: +33 (0)3 88 65 32 76
___________________________________________________________
>Hi all,
>After refining a structure with refmac, I decided to try CNS for the
>final refinement and last parametrizations (as I am more familiar with
>xplor-type parameters, than protin dictionary).
>But I have problems I get a lot of rejected reflections, even using
>sigma cutoff of zero and increase the rms Fobs (obs_rms) in order to
> not reject anything at all...
>>Why is CNS rejecting so many reflections?(it is independent of resolution)
>What should I do to include all the data?
>I am running a simple anisotropic bulk solvent correction, positional and
>bfactor.
>My data are 96% complete, but at the end I always get:
>theoretical 69728 100%
>unobserved 2555 3.7%
>reflections rejected 5047 7.2%
>total reflections used 62126 89.1%
>working set 59034 84.7%
>test set 3092 4.4%
>>Thanks in advance,
>Best regards,
>Joao Miguel
>---
---
