*** For details on how to be removed from this list visit the ***
*** CCP4 home page http://www.ccp4.ac.uk ***
Dear CCP4ers,
many thanks for your suggestions for my "too long Cd-S distance
restraint of 2.57 A problem in REFMAC" (see end of this e-mail)! Here is
a short summary:
Lothar Esser pointed to me that typical Cd-S distances in the CSD are of
the order of 2.44-2.55 A, and not 2.3 A as I have assumed. Apparently,
I've confused some ion radii: a tetrahedrally coordinated Cd2+ indeed
has a 0.15 A larger radius than a tetrahedrally coordinated Fe2+, and
thus the expected Cd-S distance is ~2.45 A (Fe-S: 2.3 A). So, the
default Cd-S distance of 2.57 A used by REFMAC is still too long, but
not as severely as I thought.
------
Eleanor Dodson recommended to set the target distances directly in
appropriate LINK cards in the PDB header like this (all on one line):
LINK SG CYS A 309 2.450 CD CD A 1
CYS-CD
This works!
------
Klaus Piontek and Sue Roberts suggested that I could modify the
$CLIBD_MON/ener_lib.cif file. I've played with the Cd ion radius without
success, and then simply copied the entry with the Fe-S target distance
to a new entry with a Cd-S target distance of 2.45 A like this:
# Cd-S copied from Fe-S and 0.15 A added
CD S metal 500.000 2.450 0.040
If I don't give any LINK cards in the PDB, the default value taken in
REFMAC is now 2.45 A (the energy factor of 500 is somewhat arbitrary, I
guess; I'm not sure whether this value is used at all; I want the TNT
libraries back!).
Anyway, this works, too!
------
Again, many thanks for your suggestions and hints!
Best wishes,
Dirk.
Dirk Kostrewa wrote:
*** For details on how to be removed from this list visit the ***
*** CCP4 home page http://www.ccp4.ac.uk ***
Dear CCP4ers,
we have a crystal structure at 2.4 A with a Cd2+ tetrahedrally
coordinated by four cysteine sulphur atoms. Irrespective of whether we
name the cadmium residue "CD " or "CD1" (and atom names "CD " or "CD1
"), REFMAC5.2 recognizes the ion as neutral cadmium atom (judged by the
total number of electrons) and reports, accordingly, too long Cd-S
dictionary distances of ~2.6 A. We expect a distance close to a Fe-S
distance of ~2.3 A. The refined structure has Cd-S distances of 2.6 A as
dictated by the dictionary . How can I convince REFMAC5.2 to use the
Cd2+ ion instead of the Cd atom (similar problem with Cl- instead of
Cl). I must admit, that after years of using REFMAC, I still don't
understand the dictionary :-(
Any help is welcome!
Best regards,
Dirk.
--
****************************************
Dirk Kostrewa
Paul Scherrer Institut
Biomolecular Research, OFLC/110
CH-5232 Villigen PSI, Switzerland
Phone: +41-56-310-4722
Fax: +41-56-310-5288
E-mail: [EMAIL PROTECTED]
http://sb.web.psi.ch
****************************************
