phenix/ccp4 monomer library compatibility
Dear all: this may have been asked a thousand times, but: are the ccp4 standard monomer libraries incompatible with phenix.refine, and if yes, is there an easy way to resolve this? I am refining a structure with an FMN, which I build with the "get monomer" option in coot. Unfortunately, phenix.refine doesn't like it and throws it out. Using a cif generated with elbow, on the other hand, has chirality errors (which I can fix...) and isn't liked by coot (0.6.2), which refuses to refine the ligand. Thanks in advance for your help, Wulf -- Prof. Dr. Wulf Blankenfeldt NW-I, 2.0.U1.09 Universitaet Bayreuth fon: +49-(0)921-55-2427 Lehrstuhl fuer Biochemie fax: +49-(0)921-55-2432 Universitaetsstrasse 30 e-mail: wulf.blankenfeldt [at] uni-bayreuth.de 95447 Bayreuth web: www.biochemie.uni-bayreuth.de Germany
On 04/09/12 12:34, Wulf Blankenfeldt wrote:
this may have been asked a thousand times, but: are the ccp4 standard monomer libraries incompatible with phenix.refine, and if yes, is there an easy way to resolve this?
I am refining a structure with an FMN, which I build with the "get monomer" option in coot. Unfortunately, phenix.refine doesn't like it and throws it out. Using a cif generated with elbow, on the other hand, has chirality errors (which I can fix...) and isn't liked by coot (0.6.2), which refuses to refine the ligand.
:-( Hi Wulf, It seem that somehow you have an inconsistency between the model and the dictionary. The correct atom names are defined by the PDB. e.g. ftp://ftp.ebi.ac.uk/pub/databases/msd/pdbechem/files/mmcif/FMN.cif When I use "Get Monomer" I get a FMN monomer this is consistent with this (well, except for phosphate hydrogens). In particular, there are primed atom names (not starred). If you are using - or referencing an old CCP4 distribution you may be picking up an old dictionary when using "Get Monomer" - coot should tell you in the console (at least) the file names that it reads. If you want to reference the Phenix copy of the monomer library, you can use use COOT_REFMAC_LIB_DIR to point to it. If you use the model (pdb file) and dictionary cif from elbow (and call it FMN) then coot should be happy no matter what the PDB says. And I would imagine that that would be so for phenix.refine also. Paul.
Speaking of dictionaries... It seems that latest Phenix is distributed with dictionaries from 2005: /usr/local/phenix-dev-1148/chem_data/mon_lib/list % head mon_lib_list.cif global_ _lib_name mon_lib _lib_version 4.11 _lib_update 15/04/05 # ------------------------------------------------ # # --- LIST OF MONOMERS --- # data_comp_list loop_ We had a problem similar to yours with some ligands. What we did about it: get latest dictionary from Garib Murshudov's group page http://www2.mrc-lmb.cam.ac.uk/groups/murshudov/ unpack rename resulting folder monomers to mon_lib fix up the beginning of mon_lib/list/mon_lib_list.cif: Line 1: underscore missing after "global" Line 6: number sign missing overwrite the existing mon_lib folder of Phenix. It seems to work. Dmitry On 2012-09-04, at 7:55 AM, Paul Emsley wrote: On 04/09/12 12:34, Wulf Blankenfeldt wrote: this may have been asked a thousand times, but: are the ccp4 standard monomer libraries incompatible with phenix.refine, and if yes, is there an easy way to resolve this? I am refining a structure with an FMN, which I build with the "get monomer" option in coot. Unfortunately, phenix.refine doesn't like it and throws it out. Using a cif generated with elbow, on the other hand, has chirality errors (which I can fix...) and isn't liked by coot (0.6.2), which refuses to refine the ligand. :-( Hi Wulf, It seem that somehow you have an inconsistency between the model and the dictionary. The correct atom names are defined by the PDB. e.g. ftp://ftp.ebi.ac.uk/pub/databases/msd/pdbechem/files/mmcif/FMN.cif When I use "Get Monomer" I get a FMN monomer this is consistent with this (well, except for phosphate hydrogens). In particular, there are primed atom names (not starred). If you are using - or referencing an old CCP4 distribution you may be picking up an old dictionary when using "Get Monomer" - coot should tell you in the console (at least) the file names that it reads. If you want to reference the Phenix copy of the monomer library, you can use use COOT_REFMAC_LIB_DIR to point to it. If you use the model (pdb file) and dictionary cif from elbow (and call it FMN) then coot should be happy no matter what the PDB says. And I would imagine that that would be so for phenix.refine also. Paul. _______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb
To clarify: CCP4 Monomer Library is a great resource. However it contains a lot of entries are not suitable (to say the least) to use in refinement. Therefore, what's phenix.refine is using is a subset of Monomer Library, that contains only usable (for refinement) entries, plus Nigel's geo standard that contains curated entries plus those complimented with some new infromation (such as R-lab rotamer library torsions). Nigel may probably provide more information about geo-standard and libraries used in Phenix in general. Pavel On 9/5/12 7:50 AM, Dmitry Rodionov wrote:
Speaking of dictionaries...
It seems that latest Phenix is distributed with dictionaries from 2005:
/usr/local/phenix-dev-1148/chem_data/mon_lib/list % head mon_lib_list.cif
global_ _lib_name mon_lib _lib_version 4.11 _lib_update 15/04/05 # ------------------------------------------------ # # --- LIST OF MONOMERS --- # data_comp_list loop_
We had a problem similar to yours with some ligands.
What we did about it:
get latest dictionary from Garib Murshudov's group page http://www2.mrc-lmb.cam.ac.uk/groups/murshudov/
unpack
rename resulting folder monomers to mon_lib
fix up the beginning of mon_lib/list/mon_lib_list.cif: Line 1: underscore missing after "global" Line 6: number sign missing
overwrite the existing mon_lib folder of Phenix.
It seems to work.
Dmitry
On 2012-09-04, at 7:55 AM, Paul Emsley wrote:
On 04/09/12 12:34, Wulf Blankenfeldt wrote:
this may have been asked a thousand times, but: are the ccp4 standard monomer libraries incompatible with phenix.refine, and if yes, is there an easy way to resolve this?
I am refining a structure with an FMN, which I build with the "get monomer" option in coot. Unfortunately, phenix.refine doesn't like it and throws it out. Using a cif generated with elbow, on the other hand, has chirality errors (which I can fix...) and isn't liked by coot (0.6.2), which refuses to refine the ligand.
:-(
Hi Wulf,
It seem that somehow you have an inconsistency between the model and the dictionary.
The correct atom names are defined by the PDB. e.g.
ftp://ftp.ebi.ac.uk/pub/databases/msd/pdbechem/files/mmcif/FMN.cif
When I use "Get Monomer" I get a FMN monomer this is consistent with this (well, except for phosphate hydrogens). In particular, there are primed atom names (not starred). If you are using - or referencing an old CCP4 distribution you may be picking up an old dictionary when using "Get Monomer" - coot should tell you in the console (at least) the file names that it reads. If you want to reference the Phenix copy of the monomer library, you can use use COOT_REFMAC_LIB_DIR to point to it.
If you use the model (pdb file) and dictionary cif from elbow (and call it FMN) then coot should be happy no matter what the PDB says. And I would imagine that that would be so for phenix.refine also.
Paul.
_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb
_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb
Ralf curated the Monomer Library (in about 2005) for suitable
restraints. Garib has updated/improved the library greatly since then
so substituting the current library is a fine idea. As Pavel said, the
GeoStd has a number of entries including the standard amino acids
which contain the rotamer information. PHENIX programs will be
supplied with the restraints from the GeoStd library if it exists,
otherwise its taken from the "Monomer Library" which ever one is in
place.
Cheers
Nigel
On Wed, Sep 5, 2012 at 9:15 AM, Pavel Afonine
To clarify:
CCP4 Monomer Library is a great resource. However it contains a lot of entries are not suitable (to say the least) to use in refinement. Therefore, what's phenix.refine is using is a subset of Monomer Library, that contains only usable (for refinement) entries, plus Nigel's geo standard that contains curated entries plus those complimented with some new infromation (such as R-lab rotamer library torsions).
Nigel may probably provide more information about geo-standard and libraries used in Phenix in general.
Pavel
On 9/5/12 7:50 AM, Dmitry Rodionov wrote:
Speaking of dictionaries...
It seems that latest Phenix is distributed with dictionaries from 2005:
/usr/local/phenix-dev-1148/chem_data/mon_lib/list % head mon_lib_list.cif
global_ _lib_name mon_lib _lib_version 4.11 _lib_update 15/04/05 # ------------------------------------------------ # # --- LIST OF MONOMERS --- # data_comp_list loop_
We had a problem similar to yours with some ligands.
What we did about it:
get latest dictionary from Garib Murshudov's group page http://www2.mrc-lmb.cam.ac.uk/groups/murshudov/
unpack
rename resulting folder monomers to mon_lib
fix up the beginning of mon_lib/list/mon_lib_list.cif: Line 1: underscore missing after "global" Line 6: number sign missing
overwrite the existing mon_lib folder of Phenix.
It seems to work.
Dmitry
On 2012-09-04, at 7:55 AM, Paul Emsley wrote:
On 04/09/12 12:34, Wulf Blankenfeldt wrote:
this may have been asked a thousand times, but: are the ccp4 standard monomer libraries incompatible with phenix.refine, and if yes, is there an easy way to resolve this?
I am refining a structure with an FMN, which I build with the "get monomer" option in coot. Unfortunately, phenix.refine doesn't like it and throws it out. Using a cif generated with elbow, on the other hand, has chirality errors (which I can fix...) and isn't liked by coot (0.6.2), which refuses to refine the ligand.
:-(
Hi Wulf,
It seem that somehow you have an inconsistency between the model and the dictionary.
The correct atom names are defined by the PDB. e.g.
ftp://ftp.ebi.ac.uk/pub/databases/msd/pdbechem/files/mmcif/FMN.cif
When I use "Get Monomer" I get a FMN monomer this is consistent with this (well, except for phosphate hydrogens). In particular, there are primed atom names (not starred). If you are using - or referencing an old CCP4 distribution you may be picking up an old dictionary when using "Get Monomer" - coot should tell you in the console (at least) the file names that it reads. If you want to reference the Phenix copy of the monomer library, you can use use COOT_REFMAC_LIB_DIR to point to it.
If you use the model (pdb file) and dictionary cif from elbow (and call it FMN) then coot should be happy no matter what the PDB says. And I would imagine that that would be so for phenix.refine also.
Paul.
_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb
_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb
_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb
-- Nigel W. Moriarty Building 64R0246B, Physical Biosciences Division Lawrence Berkeley National Laboratory Berkeley, CA 94720-8235 Phone : 510-486-5709 Email : [email protected] Fax : 510-486-5909 Web : CCI.LBL.gov
participants (5)
-
Dmitry Rodionov -
Nigel Moriarty -
Paul Emsley -
Pavel Afonine -
Wulf Blankenfeldt