Hi Juergen,
PROFESSS can be misused (?) to do this:
- just create a PDB file first the pattern atoms (numbered 1 to N)
and then the atoms of the substructure (numbered N+1 to M).
- you can then look for 'NCS' where only (1-N) atoms are related to
only (N+1-M) atoms.
Just run something like
professs xyzin pattern+substructure.pdb xysout professs.pdb <<e
DIST 50
TOLE 1.0
e
and look through professs.pdb (quite tense but it has all the
information). You might need to play with the DIST/TOLE values to
avoid PROFESSS complaining about some array limits ... or recompile
with larger limits.
And: you might need to try the other enantiomorph as well.
This worked for me a few times. I'm sure there are other possibilities
...
Cheers
Clemens
On Tue, May 27, 2008 at 03:33:19PM +0200, Juergen J. Mueller wrote:
> Dear all,
> given a very overcrowded substucture (e.g. from SHELX, PnB for sulfur)
> in a big asymmetric unit (over 100 proposed S),
> given the resolution used for the substructur determination,
> given several possible S-patterns (taken from PDB), like PAN-pattern,
> Apple-pattern a.s.o. as a pdb-file.
> Which program can be used to find the given pattern in the substructure
> (NCS search, ...) taking
> into account the error limits for the S-atom positions? Automatical
> NCS-search independent of the search pattern (using Solve/Resolve, SnB)
> does not work.
> Any hint is welcome,
> Juergen
>
--
***************************************************************
* Clemens Vonrhein, Ph.D. vonrhein AT GlobalPhasing DOT com
*
* Global Phasing Ltd.
* Sheraton House, Castle Park
* Cambridge CB3 0AX, UK
*--------------------------------------------------------------
* BUSTER Development Group (http://www.globalphasing.com)
***************************************************************
|