Dear Fred,
I think that, leaving NCS apart, you are forgetting the stereochemical
restraints on bonds and angles, which would result in lowering the number of
free parameters in the refinement to well below 20452 ... . This is an old
topic, that has been covered many times and even rather recently in this BB.
With best wishes,
Gerard.
--
On Fri, Oct 09, 2009 at 10:43:08AM +0200, Vellieux Frederic wrote:
> [log in to unmask] wrote:
>> number of atoms are 5113
>> and there are 2 chains
> Now, you have 20215 reflections used as a target to refine against (I
> assume this number does not include the Rfree reflections). You have 5113
> atoms to refine, and you refine X Y Z B-iso (4 parameters per atoms). 5113
> * 4 = 20452. Not counting the overall scale and temperature factor. Thus
> you do not have enough reflections (20215). But you have NCS. I don't think
> REFMAC can handle yet NCS (help me there Garib). CNS or Phoenix (sorry, I'm
> not American so I don't write Phenix) can handle NCS. Either as a
> constraint or as a restraint. At 3.0 Angstroem, I think you should try
> first the restraint, it should be sufficient. If you need help with
> Phoenix, you can contact Pavel Afonine, for CNS Axel Bruenger has a set of
> input template files (even in html format, to be filled using a web
> browser).
>
> Normally if you do as I suggest this should all work.
>
> HTH,
>
> Fred.
> begin:vcard
> fn:Fred. Vellieux (Ph.D)
> n:Vellieux (Ph.D);Fred.
> email;internet:[log in to unmask]
> tel;work:+33 438789605
> version:2.1
> end:vcard
>
--
===============================================================
* *
* Gerard Bricogne [log in to unmask] *
* *
* Global Phasing Ltd. *
* Sheraton House, Castle Park Tel: +44-(0)1223-353033 *
* Cambridge CB3 0AX, UK Fax: +44-(0)1223-366889 *
* *
===============================================================
|