Re: [ccp4bb] Refining alternate ligand/autoinhibition

[Stefan Arold]

Thanks to Eleanor and Boaz for their quick reply;
Here some additional infos:
I am using ccp4 6.0; Refmac_5.2.0019version 5.2.0019
and the data are only to about 2.8 A

I have tried to leave out the alternate flags - still the same: the 
repulsion persists...
I have also tried to use the same chain label for the peptide and the 
protein (to simulate two conformations of a pseudo-hybrid unit), but 
this didn't help either...


best wishes
Stefan







Eleanor Dodson a écrit :

Well - A and B are meant to be flags to show there are 2 conformations 
of a particular unit. sp they arent appropriate here and may be 
muddling things up.


On the other hand you shouldnt get any repulsion between pairs of 
atoms whose added occupancy is <= to 1.0


Is that always the case for your 2 scenarios?

Try removing the A and B flags and see if that is better.
Eleanor


Stefan Arold wrote:


Dear All,

I am having trouble refining a difficult case of alternate structures 
in REFMAC. In the crystal, two states superimpose:
1) Apo-protein with its own auto-inhibitory tail fixed into the 
ligand binding site

and
2) Peptide-bound protein, where the peptide has displaced the 
auto-inhibitory tail.
this also includes alternate side-chains of the protein ligand 
binding site, according to the apo state or peptide-bound state


I was hoping that assigning 'A' and 'B' alternate states (for 'apo' 
and 'peptide-bound') to the protein chain (chain A)

[ex:
ATOM 62  C  ATYR A  17  20.856  42.150  21.476  0.60 
75.23   C]


and state 'B' to the peptide
[ex:
ATOM 87  N  BALA D 513   9.910  69.145  50.780  0.40 
20.00   N]


would do the job - but unfortunately, even though refmac refines this 
structure, the two states 'sense' each other, and refmac refinement 
applies repulsive forces, pulling the apo-auto-inhibitory tail and 
the peptide away from each other...


Neither the refmac manual, nor the ccp4bb archives helped, so I 
thought one of you might...


many thanks

Stefan








--
==
Stefan T. Arold, PhD
Centre de Biochimie Structurale
CNRS UMR 5048 - UM 1 - INSERM UMR 554
29 rue de Navacelles 34090
MONTPELLIER Cedex  - France
email: [EMAIL PROTECTED]
Phone: +33 (0)4.67.41.77.02
Fax:   +33 (0)4.67.41.79.13
==

Re: [ccp4bb] Refining alternate ligand/autoinhibition

[Eleanor Dodson]

Well - A and B are meant to be flags to show there are 2 conformations 
of a particular unit. sp they arent appropriate here and may be muddling 
things up.


On the other hand you shouldnt get any repulsion between pairs of atoms 
whose added occupancy is <= to 1.0


Is that always the case for your 2 scenarios?

Try removing the A and B flags and see if that is better.
Eleanor


Stefan Arold wrote:

Dear All,

I am having trouble refining a difficult case of alternate structures 
in REFMAC. In the crystal, two states superimpose:
1) Apo-protein with its own auto-inhibitory tail fixed into the ligand 
binding site

and
2) Peptide-bound protein, where the peptide has displaced the 
auto-inhibitory tail.
this also includes alternate side-chains of the protein ligand binding 
site, according to the apo state or peptide-bound state


I was hoping that assigning 'A' and 'B' alternate states (for 'apo' 
and 'peptide-bound') to the protein chain (chain A)

[ex:
ATOM 62  C  ATYR A  17  20.856  42.150  21.476  0.60 
75.23   C]


and state 'B' to the peptide
[ex:
ATOM 87  N  BALA D 513   9.910  69.145  50.780  0.40 
20.00   N]


would do the job - but unfortunately, even though refmac refines this 
structure, the two states 'sense' each other, and refmac refinement 
applies repulsive forces, pulling the apo-auto-inhibitory tail and the 
peptide away from each other...


Neither the refmac manual, nor the ccp4bb archives helped, so I 
thought one of you might...


many thanks

Stefan