Re: [ccp4bb] Modelling of ligand and Refmac5

[Sam Tang]

Dear Dr Emsley

Many thanks for your reply.

I have now updated my Coot and am able to run refmac refinement using my
ligand PDB and CIF generated from elbow2 under Phenix.

This is indeed a lesson on the importance of keeping all softwares updated!

Kind regards

Sam

On 24 January 2017 at 18:29, Paul Emsley  wrote:

> On 24/01/2017 08:04, Sam Tang wrote:
>
>>
>> I am trying to fit a small molecule ligand into a protein complex using
>> Coot. The data was
>> processed to P212121, at 2.6 A.  What I did was to input a SMILES string,
>> fit the ligand,
>> merge the ligand into the protein molecule using 'Merge Molecules' and
>> save coordinates.
>>
>> After fitting the ligand (now as chain O) I ran restrained refinement in
>> Refmac5 and the
>> following error returns:
>>
>>
> Into what did you input a SMILES? If the answer is Coot, then you have an
> Old Coot and will be lead down the garden path.
>
> > I believe it is due to the nomenclature of the ligand wherein Refmac
> mistook atoms as DUM
>
> I think that you're more or less right.
>
> The modern approach is to use Acedrg to generate the ligand either via a
> GUI or the command line. That will give you a PDB file which you can fit,
> and a dictionary that you can use in Refmac, Coot (and, I believe, Phenix).
>
> Paul.
>

Re: [ccp4bb] Modelling of ligand and Refmac5

[Paul Emsley]

On 24/01/2017 08:04, Sam Tang wrote:


I am trying to fit a small molecule ligand into a protein complex using Coot. 
The data was
processed to P212121, at 2.6 A.  What I did was to input a SMILES string, fit 
the ligand,
merge the ligand into the protein molecule using 'Merge Molecules' and save 
coordinates.

After fitting the ligand (now as chain O) I ran restrained refinement in 
Refmac5 and the
following error returns:



Into what did you input a SMILES? If the answer is Coot, then you have an Old Coot and will 
be lead down the garden path.


> I believe it is due to the nomenclature of the ligand wherein Refmac mistook 
atoms as DUM

I think that you're more or less right.

The modern approach is to use Acedrg to generate the ligand either via a GUI or the command 
line. That will give you a PDB file which you can fit, and a dictionary that you can use in 
Refmac, Coot (and, I believe, Phenix).


Paul.

[ccp4bb] Modelling of ligand and Refmac5

[Sam Tang]

Dear all

I am trying to fit a small molecule ligand into a protein complex using
Coot. The data was processed to P212121, at 2.6 A.  What I did was to input
a SMILES string, fit the ligand, merge the ligand into the protein molecule
using 'Merge Molecules' and save coordinates.

After fitting the ligand (now as chain O) I ran restrained refinement in
Refmac5 and the following error returns:

Input file :C:/20161226_1.1_refmac1-coot-1.pdb

  --

  ---  LIBRARY OF MONOMERS   ---

 _lib_name mon_lib

 _lib_version  5.44

 _lib_update   30/05/14

  --

  NUMBER OF MONOMERS IN THE LIBRARY  : 13409

with complete description: 13409

  NUMBER OF MODIFICATIONS:63

  NUMBER OF LINKS:73

  I am reading libraries. Please wait.

  - energy parameters

  - monomer"s description (links & mod )


FORMATTED  OLD file opened on unit  45



Logical name: ATOMSF, Filename: C:\CCP4-7\7.0\lib\data\atomsf.lib




  Number of atoms:   19934

  Number of residues :2577

  Number of chains   :  15

  I am reading library. Please wait.

mon_lib.cif


ERROR : DUM : duplicated atom_name : "DUM ".

  chain: OO   residue:1


(And the error repeats itself for >100 times)


I believe it is due to the nomenclature of the ligand wherein Refmac
mistook atoms as DUM.  (Also, the chain ID O was identified as OO?)  In a
test run I carried out rigid body refinement and the programme finished
without issues.

Is there a way I could rectify the above problem? Thanks in advance for
your attention and input.

Kind regards


Sam Tang
Biochemistry Programme, School of Life Sciences, CUHK