Error with Phenix refine after adding ligand
Hello all, I run into this problem when I tried to refine my protein structure after adding ligand. The ligand that I am adding is a malonate, and the 3-letter chemical code is MLI. I add the ligand using Coot and it fit perfectly into the electron density map. I then tried to run phenix.refine by giving the protein structure pdb file, the original mtz file and the protein sequence fasta file. However, there was an error message saying missing restraints for the ligand. So I generate the CIF file for the ligand using the Ready Set icon in the toolbar. The ligand CIF file and a updated protein structure pdb file was generated and automatically added to the refinement windows. However, when I run the refinement, I saw this error message: Number of nonbonded interaction distances <0.001:2. Please inspect the output above (or in the log window) and correct the input model file. When I inspect the log window, the error is due to the oxygen atoms in the ligand being too close to each other. Why is there an error if the ligand was generated automatically in Phenix? Has anyone encounter this problem before? How do you solve that? Many thanks in advance! -- Dr Mian-Chee Gor Research Fellow Griffith Institute for Drug Discovery (GRIDD) Don Young Road, Nathan, QLD 4111 Mobile: +61 4 2084 7785 Email: [email protected]
Hi, I can have a closer look into this issue if : a) you send me all relevant input files, and b) describe what you did (step-by-step) so that I try to reproduce your steps up to the crash and then see how we can address this. If you choose to send files, please make sure to send them to my email address, not the entire mailing list! Pavel On 1/22/18 21:20, Mian-Chee Gor wrote:
Hello all,
I run into this problem when I tried to refine my protein structure after adding ligand.
The ligand that I am adding is a malonate, and the 3-letter chemical code is MLI. I add the ligand using Coot and it fit perfectly into the electron density map.
I then tried to run phenix.refine by giving the protein structure pdb file, the original mtz file and the protein sequence fasta file. However, there was an error message saying missing restraints for the ligand. So I generate the CIF file for the ligand using the Ready Set icon in the toolbar. The ligand CIF file and a updated protein structure pdb file was generated and automatically added to the refinement windows. However, when I run the refinement, I saw this error message: Number of nonbonded interaction distances <0.001:2. Please inspect the output above (or in the log window) and correct the input model file.
When I inspect the log window, the error is due to the oxygen atoms in the ligand being too close to each other. Why is there an error if the ligand was generated automatically in Phenix? Has anyone encounter this problem before? How do you solve that?
Many thanks in advance!
-- Dr Mian-Chee Gor Research Fellow Griffith Institute for Drug Discovery (GRIDD) Don Young Road, Nathan, QLD 4111 Mobile: +61 4 2084 7785 Email: [email protected] mailto:[email protected]
_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb Unsubscribe: [email protected]
Hi Main-Chee,
The ligand that I am adding is a malonate, and the 3-letter chemical code is MLI. I add the ligand using Coot and it fit perfectly into the electron density map.
I suspect this is when you ended up with a model with two oxygens sitting on top of each other. Since you already identified oxygens in question it would be easy to check if Coot introduced them before you run any Phenix program. In this case I would address this question to Coot developers. Best regards, Oleg Sobolev.
participants (3)
-
Mian-Chee Gor -
Oleg Sobolev -
Pavel Afonine