Print

Print
Hi Greg, Herman all others in CCP4 who can help,
Thank you for your suggestions which I followed, but I am still not able to make the link between the F3P molecule and the first residue of protein chain GLycine.
Here is how I have modified the PDB after merging the molecules.

CRYST1   16.950   44.250   66.410  90.00  90.00  90.00 P 21 21 21               
SCALE1      0.058997  0.000000  0.000000        0.00000                         
SCALE2      0.000000  0.022599  0.000000        0.00000                         
SCALE3      0.000000  0.000000  0.015058        0.00000                         
HETATM  585  C1  F3P A   0      -3.340  -3.808  27.614  1.00 20.00           C  
HETATM  586  C2  F3P A   0      -3.613  -2.393  27.654  1.00 20.00           C  
HETATM  587  F   F3P A   0      -2.442  -1.672  27.870  1.00 20.00           F  
HETATM  588  C3  F3P A   0      -4.477  -1.785  26.669  1.00 20.00           C  
HETATM  589  C10 F3P A   0      -3.926  -0.760  25.891  1.00 20.00           C  
HETATM  590  O1  F3P A   0      -5.640  -2.341  26.364  1.00 20.00           O  
ATOM      1  N   GLY A   1      -3.846  -0.932  25.879  1.00  5.57           N  
ATOM      2  CA  GLY A   1      -4.553  -0.083  24.938  1.00  5.14           C  
ATOM      3  C   GLY A   1      -4.433  -0.577  23.509  1.00  4.49           C  
ATOM      4  O   GLY A   1      -3.832  -1.619  23.250  1.00  5.72           O  


I have attached the .cif file as well as the PDB of the F3P molecule as well as the images that I am getting when I am trying to  link the two molecules as suggested my Herman and Greg.
We do not have something similar in PDB unfortunately.

As you have noticed the molecule diffracted to 0.95 Angs and I got several datasets at 1.02-1.05 Angs and the density is clean.All the amino acid residues are built like a charm.
Yet it is taking me ages to complete this structure!!!
I thought there should be a separates protocol describing the steps in CCP4/COOT on how to do this type of covalent linking on any molecule (as many people are trying to build new molecules (which are not always found in the PDB with such restraint information) that so many protein molecules are modified for drugs and other purposes.
Once I get a solution, I will write up everything and then the CCP4/COOT developers could edit and use it as an addendum in the "How to".

Thanks again in advance for your suggestions and the time you are taking to read through the problem. I sincerely appreciate it.
Ivan

On Thu, Jul 16, 2015 at 11:36 PM, <[log in to unmask]> wrote:

Hi Ivan,

 

Giving your “ligand” the same residue number and chain ID as the first residue of your protein definitively will cause problems like being unable to refine it. There is nothing against giving the “ligand” residue number 0 (zero) what I would do.

Also, if I understand correctly, your “ligand” is covalently linked to the N-terminus. In this case you have to specify the link in your pdb file. In coot you can do this by clicking 2 atoms (Extensions-Modelling-Make Link). This has to be done after the “ligand” and protein have been merged. You also have to provide a cif file describing this link. This is non-trivial, but should all be described in the Refmac documentation. I would also check whether this modification already exist in the pdb, so you could use existing definitions.

 

Good luck!

Herman

 

 

 

Von: CCP4 bulletin board [mailto:[log in to unmask]] Im Auftrag von xaravich ivan
Gesendet: Freitag, 17. Juli 2015 04:59
An: [log in to unmask]
Betreff: Re: [ccp4bb] Coot Smiles would not build

 

Hi everyone,

I am definitely on my way to becoming crazy!!! I took the pdb created by Greg and I prepared my own .cif file in CCP4  as well as used the .cif file prepared by Greg but once i merge the molecule after placing it in the density. it somehow doesnot refine. I merge the restraints into one cif file and still it does not let me refine even in COOT when I am using Calculate>Model/fit/refine>Real space refine zone, right after I place the molecule. I am deleting the hydrogens though before I place the molecule into the density (before refinement image). After running Refmac5 I lose most of the density as you see in the image and the molecule is misplaced.

Another problem is that this molecule is basically placed at the N-terminal of residue 1 of chainA so it also has the number 1. I do not know if this is creating the problem. I chnaged the chain name into Chain A, as this is a residue in chain A and renumbered it to the last residue of the peptide instead of renumbering all the aminoacids and still no luck.

 

What should I do, I am running out of ideas. 

Am I doing the steps right?

1) I generate or get a pdb of the ligand molecule (basically it is not ligand in that sense!!)

2) Generate restraints by ProDrg

3) Open the pdb of the molecule, and import cif file and then open the PDB protein molecule and the corresponding map

4) Fiting the ligand molecule into the protein Map

5) Merge molecule in COOT

 

I can't explain how much I would appreciate your help and suggestions.

 

Thanks,

Ivan

 

 

On Wed, Jul 15, 2015 at 9:41 AM, Paul Emsley <[log in to unmask]> wrote:

On 11/07/2015 01:02, xaravich ivan wrote:

Hi everyone
I am using SMILES to build this Fluopropyl group and SMILES would always give me this error, after I put the correct string.
The image and string I built and the error image is attached.

Am I missing something?

For the record, the problem occurred because you were running coot in a directory in which it was unable to write files and directories. (This is not recommended.)

Paul.