Hi Xiao,
 
 
You need a cif file for the IAS, because IAS is not included in the standard monomer lib. Please see the attached IAS.cif I made from ASP.cif.
 
Compared to ASP.cif, this file has the following changes:
1) compound 3-letter code  to IAS,
2) compound name            to 'isoASPARTIC-ACID                       '
3) _chem_comp.group       to L-aminoacid (so that coot won’t think it is part of a standard L-peptide and try to link next aa to the C atom. We want the CG to be linked to next N now.)
 
The file seems to work for coot (you should be able to drag the IAS around when you have done “import CIF dictionary” with it). Hopefully it will work for refmac and phenix.refine too .
 
 
Zhijie
 
 
 
From: Xiao Lei
Sent: Tuesday, August 18, 2015 1:56 PM
Subject: Re: [phenixbb] Coot mutation of Asp residue to isoAsp residue
 
Hi Zhejie,
 
For the 1AT6 structure, I downloaded its density in coot using "fetch density from EDS", but when I found the IAS at 101 position and try to do real space refine, it gives an error that "Refinement setup failure. Failed to find restraints for IAS."
 
I do not know how to fix this but it seems to me it's caused by incomplete restraints dictionary or monomer library in ccp4?
 
Thanks.
 
Xiao
 
On Tue, Aug 18, 2015 at 10:42 AM, Xiao Lei <[email protected]> wrote:
Hi Zhijie,
 
Thank you very much for the information. For step 1 you mentioned, I can get monomer with L-Asp but it seems I can not drag it (or I do not know how to do) and can not delete or modify it to become isoAsp. I will try play around more though.
 
Xiao
 
On Mon, Aug 17, 2015 at 6:33 PM, Zhijie Li <[email protected]> wrote:
Hi Xiao,
 
IsoAsp is essentially an L-Asp linked with next aa through its side chain (beta) carboxyl. So the mutation button won’t help you. You need to build in a new L-ASP, which is treated as a covalently linked ligand (HETATM records), instead of a standard residue (ATOM records) of the protein chain.
 
A practical method might be: 1) delete the original Asp, 2) import a free L-Asp using “get monomer”, delete its hydrogen atoms and drag it into the density, delete one oxygen atom on the beta-carboxyl and change the residue’s numbering and chain id to fit it into the sequence,  3) edit the PDB, if necessary, to turn the ASP into a ligand (a HETATM record inside the chain).
 
For step 3, you may need to rename the ASP to something else (IAS was used for isoASP in older pdb, so I would go with IAS ) so that coot won’t try to make a regular peptide bond using its main chain carboxyl during real space refinement. Of course you will need to make a cif file for the “new” compound too. I guess you can make a copy of ASP.cif from the monomer library and change everything in it to IAS. I think if you have placed the IAS to the right location and its ends are in bonding distance with the neighbouring aa residues you may not need to do anything for refmac. For phenix.refine you will need to add a bond description to the .edit file for each linkage the IAS makes to the neighboring aas.
 
You may take a look at the structure 1AT6 and its PDB file. The residue IAS 101 is an example of isoASP. Note that the IAS atoms are HETATM in the chain and there are two LINK records in the header to indicate its linkage to neighbouring aas (LINK records are normally not generated or needed during refinement using refmac or phenix.refine).   
 
 
Zhijie
 
 
 
From: Xiao Lei
Sent: Monday, August 17, 2015 6:50 PM
Subject: [phenixbb] Coot mutation of Asp residue to isoAsp residue
 
Dear Phenixbb members,
 
I suspect one Asp residue in my model may be an isoAsp (isomerization of Asp). I am asking if there is way to mutate Asp residue to isoAsp(isoaspartic acid) residue in coot GUI (I'm using coot 0.8.1 EL in Mac OS X10.10.5)?
 
I know there is a mutation button on coot, but the mutated aa lists are all natural amino acids. If I have to delete the Asp residue first and then build isoAsp into the density map, is there a way in coot to build an isoAsp residue in map?
 
Thanks ahead.
 
Xiao
_______________________________________________
phenixbb mailing list
[email protected]
http://phenix-online.org/mailman/listinfo/phenixbb
Unsubscribe: [email protected]