Obtaining standard cif restrain file for ligand in PDB
Dear all, I am building a model of a digitonin molecule into the density map. Several released PDB entries contain digitonin (ligand code: AJP). However, I could not find the corresponding mmCIF restraint file (AJP.cif) for refinement. I generated one using phenix.elbow, but it failed PDB validation, likely because the generated parameters differ from the standard reference. Could you please advise on where I can obtain the standard restraint file for AJP? I have checked the RCSB PDB website, but it appears to provide coordinate files rather than refinement restraints. Thank you for your help. Best regards, Lei
Dear Lei, It sounds to me that you did everything alright (but I didn't see the refined structure obviously). From my experience, there are still some limitations in ligand validation in the PDB deposition. So if you are sure that your restraint file is good and refinement result makes sense, then do not worry much about reported outliers in the PDB deposition. (This may be a bit controversial opinion and someone can disagree.) (I should note that the PDB team is making a lot of effort to improve the ligand validation and is getting better and better. It's very appreciated. So may change my controversial opinion in the near future.) You can also try to use a restraint file from the CCP4 monomer library which was generated in AceDRG: https://raw.githubusercontent.com/MonomerLibrary/monomers/refs/heads/master/... Or you can generate a restraint file using GRADE server of Global Phasing. Then you can check how consistent are you refinement results while using restraints generated in different programs - if you had any doubts. Best wishes, Martin On 16/01/2026 08:32, Lei Chen wrote:
Dear all,
I am building a model of a digitonin molecule into the density map. Several released PDB entries contain digitonin (ligand code: AJP).
However, I could not find the corresponding mmCIF restraint file (AJP.cif) for refinement. I generated one using phenix.elbow, but it failed PDB validation, likely because the generated parameters differ from the standard reference.
Could you please advise on where I can obtain the standard restraint file for AJP? I have checked the RCSB PDB website, but it appears to provide coordinate files rather than refinement restraints.
Thank you for your help.
Best regards, Lei _______________________________________________ phenixbb mailing list -- [email protected] To unsubscribe send an email to [email protected] Unsubscribe: phenixbb-leave@%(host_name)s
Dear Martin, This AceDRG resource looks very useful. Is there a way to search this online monomer library? Or can I download the whole library? Thanks, Lei On 1/16/2026 5:26 PM, Martin Malý wrote:
Dear Lei,
It sounds to me that you did everything alright (but I didn't see the refined structure obviously). From my experience, there are still some limitations in ligand validation in the PDB deposition. So if you are sure that your restraint file is good and refinement result makes sense, then do not worry much about reported outliers in the PDB deposition. (This may be a bit controversial opinion and someone can disagree.) (I should note that the PDB team is making a lot of effort to improve the ligand validation and is getting better and better. It's very appreciated. So may change my controversial opinion in the near future.)
You can also try to use a restraint file from the CCP4 monomer library which was generated in AceDRG: https://raw.githubusercontent.com/MonomerLibrary/monomers/refs/heads/master/...
Or you can generate a restraint file using GRADE server of Global Phasing. Then you can check how consistent are you refinement results while using restraints generated in different programs - if you had any doubts.
Best wishes, Martin
On 16/01/2026 08:32, Lei Chen wrote:
Dear all,
I am building a model of a digitonin molecule into the density map. Several released PDB entries contain digitonin (ligand code: AJP).
However, I could not find the corresponding mmCIF restraint file (AJP.cif) for refinement. I generated one using phenix.elbow, but it failed PDB validation, likely because the generated parameters differ from the standard reference.
Could you please advise on where I can obtain the standard restraint file for AJP? I have checked the RCSB PDB website, but it appears to provide coordinate files rather than refinement restraints.
Thank you for your help.
Best regards, Lei _______________________________________________ phenixbb mailing list -- [email protected] To unsubscribe send an email to [email protected] Unsubscribe: phenixbb-leave@%(host_name)s
Hello,
You can at least browse the repository: https://github.com/MonomerLibrary/monomers
Probably even search it, but it seems this only works when logged in with a GitHub account (can’t check now on a phone).
You can also download the entire repository with git clone and update your copy later with git pull. This gives you direct access to the cif files.
Cheers,
Guillaume
Typed on a phone
On 17 Jan 2026, at 06:06, rainfieldcn
Hi Lei,
Not sure if you already tried this but will it possible to use the ideal sdf file to make the restraint file using elbow
Abhi
From: Lei Chen
Hi Lei, eLBOW or ReadySet! are the tools for this, so you are on the right track. Failing PDB validation can mean two things, with a 50/50 chance: Phenix is wrong or PDB is wrong. So here, you are on the wrong track by blindly assuming Phenix is wrong. The command mmtbx.where_is_that_cif_file AJP shows that Phenix has AJP in its library. This means you do not even need to generate restraints for it using eLBOW or ReadySet! If you have correctly defined AJP in your model file (correctly means atom names conform with the library), you should be golden. Good luck! Pavel On 1/16/26 00:32, Lei Chen wrote:
Dear all,
I am building a model of a digitonin molecule into the density map. Several released PDB entries contain digitonin (ligand code: AJP).
However, I could not find the corresponding mmCIF restraint file (AJP.cif) for refinement. I generated one using phenix.elbow, but it failed PDB validation, likely because the generated parameters differ from the standard reference.
Could you please advise on where I can obtain the standard restraint file for AJP? I have checked the RCSB PDB website, but it appears to provide coordinate files rather than refinement restraints.
Thank you for your help.
Best regards, Lei _______________________________________________ phenixbb mailing list -- [email protected] To unsubscribe send an email to [email protected] Unsubscribe: phenixbb-leave@%(host_name)s
You can also send the files directly to me so I can investigate. AJP has a
38 chiral restraints requiring a match between the input model and the
restraints.
Cheers
Nigel
---
Nigel W. Moriarty
Building 91, Molecular Biophysics and Integrated Bioimaging
Lawrence Berkeley National Laboratory
Berkeley, CA 94720-8235
Email : [email protected]
Web : CCI.LBL.gov
ORCID : orcid.org/0000-0001-8857-9464
On Fri, Jan 16, 2026 at 8:53 AM Pavel Afonine
Hi Lei,
eLBOW or ReadySet! are the tools for this, so you are on the right track.
Failing PDB validation can mean two things, with a 50/50 chance: Phenix is wrong or PDB is wrong. So here, you are on the wrong track by blindly assuming Phenix is wrong.
The command
mmtbx.where_is_that_cif_file AJP
shows that Phenix has AJP in its library. This means you do not even need to generate restraints for it using eLBOW or ReadySet! If you have correctly defined AJP in your model file (correctly means atom names conform with the library), you should be golden.
Good luck! Pavel
On 1/16/26 00:32, Lei Chen wrote:
Dear all,
I am building a model of a digitonin molecule into the density map. Several released PDB entries contain digitonin (ligand code: AJP).
However, I could not find the corresponding mmCIF restraint file (AJP.cif) for refinement. I generated one using phenix.elbow, but it failed PDB validation, likely because the generated parameters differ from the standard reference.
Could you please advise on where I can obtain the standard restraint file for AJP? I have checked the RCSB PDB website, but it appears to provide coordinate files rather than refinement restraints.
Thank you for your help.
Best regards, Lei _______________________________________________ phenixbb mailing list -- [email protected] To unsubscribe send an email to [email protected] Unsubscribe: phenixbb-leave@%(host_name)s
phenixbb mailing list -- [email protected] To unsubscribe send an email to [email protected] Unsubscribe: phenixbb-leave@%(host_name)s
The error I got was about bond length. The chirals are OK.
Thank you though.
Best,
Lei
On Sat, Jan 17, 2026 at 2:17 AM Nigel Moriarty
You can also send the files directly to me so I can investigate. AJP has a 38 chiral restraints requiring a match between the input model and the restraints.
Cheers
Nigel
--- Nigel W. Moriarty Building 91, Molecular Biophysics and Integrated Bioimaging Lawrence Berkeley National Laboratory Berkeley, CA 94720-8235 Email : [email protected] Web : CCI.LBL.gov ORCID : orcid.org/0000-0001-8857-9464
On Fri, Jan 16, 2026 at 8:53 AM Pavel Afonine
wrote: Hi Lei,
eLBOW or ReadySet! are the tools for this, so you are on the right track.
Failing PDB validation can mean two things, with a 50/50 chance: Phenix is wrong or PDB is wrong. So here, you are on the wrong track by blindly assuming Phenix is wrong.
The command
mmtbx.where_is_that_cif_file AJP
shows that Phenix has AJP in its library. This means you do not even need to generate restraints for it using eLBOW or ReadySet! If you have correctly defined AJP in your model file (correctly means atom names conform with the library), you should be golden.
Good luck! Pavel
On 1/16/26 00:32, Lei Chen wrote:
Dear all,
I am building a model of a digitonin molecule into the density map. Several released PDB entries contain digitonin (ligand code: AJP).
However, I could not find the corresponding mmCIF restraint file (AJP.cif) for refinement. I generated one using phenix.elbow, but it failed PDB validation, likely because the generated parameters differ from the standard reference.
Could you please advise on where I can obtain the standard restraint file for AJP? I have checked the RCSB PDB website, but it appears to provide coordinate files rather than refinement restraints.
Thank you for your help.
Best regards, Lei _______________________________________________ phenixbb mailing list -- [email protected] To unsubscribe send an email to [email protected] Unsubscribe: phenixbb-leave@%(host_name)s
phenixbb mailing list -- [email protected] To unsubscribe send an email to [email protected] Unsubscribe: phenixbb-leave@%(host_name)s
Lei
As the author of both eLBOW and ReadySet!, I'm sure I can help but I need
more. If you wish to share the model and restraints (directly and
privately) I can take a look otherwise there is nothing anyone can do with
the information you have provided so far.
Cheers
Nigel
---
Nigel W. Moriarty
Building 91, Molecular Biophysics and Integrated Bioimaging
Lawrence Berkeley National Laboratory
Berkeley, CA 94720-8235
Email : [email protected]
Web : CCI.LBL.gov
ORCID : orcid.org/0000-0001-8857-9464
On Fri, Jan 16, 2026 at 10:15 PM Lei Chen
The error I got was about bond length. The chirals are OK. Thank you though. Best, Lei
On Sat, Jan 17, 2026 at 2:17 AM Nigel Moriarty
wrote: You can also send the files directly to me so I can investigate. AJP has
a 38 chiral restraints requiring a match between the input model and the restraints.
Cheers
Nigel
--- Nigel W. Moriarty Building 91, Molecular Biophysics and Integrated Bioimaging Lawrence Berkeley National Laboratory Berkeley, CA 94720-8235 Email : [email protected] Web : CCI.LBL.gov ORCID : orcid.org/0000-0001-8857-9464
On Fri, Jan 16, 2026 at 8:53 AM Pavel Afonine
wrote: Hi Lei,
eLBOW or ReadySet! are the tools for this, so you are on the right
track.
Failing PDB validation can mean two things, with a 50/50 chance: Phenix is wrong or PDB is wrong. So here, you are on the wrong track by blindly assuming Phenix is wrong.
The command
mmtbx.where_is_that_cif_file AJP
shows that Phenix has AJP in its library. This means you do not even need to generate restraints for it using eLBOW or ReadySet! If you have correctly defined AJP in your model file (correctly means atom names conform with the library), you should be golden.
Good luck! Pavel
On 1/16/26 00:32, Lei Chen wrote:
Dear all,
I am building a model of a digitonin molecule into the density map. Several released PDB entries contain digitonin (ligand code: AJP).
However, I could not find the corresponding mmCIF restraint file (AJP.cif) for refinement. I generated one using phenix.elbow, but it failed PDB validation, likely because the generated parameters differ from the standard reference.
Could you please advise on where I can obtain the standard restraint file for AJP? I have checked the RCSB PDB website, but it appears to provide coordinate files rather than refinement restraints.
Thank you for your help.
Best regards, Lei _______________________________________________ phenixbb mailing list -- [email protected] To unsubscribe send an email to [email protected] Unsubscribe: phenixbb-leave@%(host_name)s
phenixbb mailing list -- [email protected] To unsubscribe send an email to [email protected] Unsubscribe: phenixbb-leave@%(host_name)s
Dear Nigel,
I resolved this issue by manually editing the CIF restraint file.
The problem occurred when I used an old version of phenix.elbow
(1.19.2-4158) to generate the CIF file for ligand AJP. The resulting
file contained incorrect geometry; for example, the C05-C06 bond
distance was listed as 1.75 Å instead of the expected ~1.53 Å. The
refinement succeeded after I corrected these values manually.
Best,
Lei
On Sun, Jan 18, 2026 at 2:14 AM Nigel Moriarty
Lei
As the author of both eLBOW and ReadySet!, I'm sure I can help but I need more. If you wish to share the model and restraints (directly and privately) I can take a look otherwise there is nothing anyone can do with the information you have provided so far.
Cheers
Nigel
--- Nigel W. Moriarty Building 91, Molecular Biophysics and Integrated Bioimaging Lawrence Berkeley National Laboratory Berkeley, CA 94720-8235 Email : [email protected] Web : CCI.LBL.gov ORCID : orcid.org/0000-0001-8857-9464
On Fri, Jan 16, 2026 at 10:15 PM Lei Chen
wrote: The error I got was about bond length. The chirals are OK. Thank you though. Best, Lei
On Sat, Jan 17, 2026 at 2:17 AM Nigel Moriarty
wrote: You can also send the files directly to me so I can investigate. AJP has a 38 chiral restraints requiring a match between the input model and the restraints.
Cheers
Nigel
--- Nigel W. Moriarty Building 91, Molecular Biophysics and Integrated Bioimaging Lawrence Berkeley National Laboratory Berkeley, CA 94720-8235 Email : [email protected] Web : CCI.LBL.gov ORCID : orcid.org/0000-0001-8857-9464
On Fri, Jan 16, 2026 at 8:53 AM Pavel Afonine
wrote: Hi Lei,
eLBOW or ReadySet! are the tools for this, so you are on the right track.
Failing PDB validation can mean two things, with a 50/50 chance: Phenix is wrong or PDB is wrong. So here, you are on the wrong track by blindly assuming Phenix is wrong.
The command
mmtbx.where_is_that_cif_file AJP
shows that Phenix has AJP in its library. This means you do not even need to generate restraints for it using eLBOW or ReadySet! If you have correctly defined AJP in your model file (correctly means atom names conform with the library), you should be golden.
Good luck! Pavel
On 1/16/26 00:32, Lei Chen wrote:
Dear all,
I am building a model of a digitonin molecule into the density map. Several released PDB entries contain digitonin (ligand code: AJP).
However, I could not find the corresponding mmCIF restraint file (AJP.cif) for refinement. I generated one using phenix.elbow, but it failed PDB validation, likely because the generated parameters differ from the standard reference.
Could you please advise on where I can obtain the standard restraint file for AJP? I have checked the RCSB PDB website, but it appears to provide coordinate files rather than refinement restraints.
Thank you for your help.
Best regards, Lei _______________________________________________ phenixbb mailing list -- [email protected] To unsubscribe send an email to [email protected] Unsubscribe: phenixbb-leave@%(host_name)s
phenixbb mailing list -- [email protected] To unsubscribe send an email to [email protected] Unsubscribe: phenixbb-leave@%(host_name)s
I probably used an outdated version of PHENIX and could not find the
AJP.cif file in the PHENIX/modules/chem_data/mon_lib/ directory. I'll
try getting a new version.
Thanks,
Lei
On Sat, Jan 17, 2026 at 12:53 AM Pavel Afonine
Hi Lei,
eLBOW or ReadySet! are the tools for this, so you are on the right track.
Failing PDB validation can mean two things, with a 50/50 chance: Phenix is wrong or PDB is wrong. So here, you are on the wrong track by blindly assuming Phenix is wrong.
The command
mmtbx.where_is_that_cif_file AJP
shows that Phenix has AJP in its library. This means you do not even need to generate restraints for it using eLBOW or ReadySet! If you have correctly defined AJP in your model file (correctly means atom names conform with the library), you should be golden.
Good luck! Pavel
On 1/16/26 00:32, Lei Chen wrote:
Dear all,
I am building a model of a digitonin molecule into the density map. Several released PDB entries contain digitonin (ligand code: AJP).
However, I could not find the corresponding mmCIF restraint file (AJP.cif) for refinement. I generated one using phenix.elbow, but it failed PDB validation, likely because the generated parameters differ from the standard reference.
Could you please advise on where I can obtain the standard restraint file for AJP? I have checked the RCSB PDB website, but it appears to provide coordinate files rather than refinement restraints.
Thank you for your help.
Best regards, Lei _______________________________________________ phenixbb mailing list -- [email protected] To unsubscribe send an email to [email protected] Unsubscribe: phenixbb-leave@%(host_name)s
participants (7)
-
Das, Abhinaba -
Guillaume Gaullier -
Lei Chen -
Martin Malý -
Nigel Moriarty -
Pavel Afonine -
rainfieldcn