[Dock-fans] FW: prepare_amber.pl

Rahul Banerjee rahul.banerjee at chemistry.msu.edu
Tue Jun 30 14:02:19 PDT 2009


Dear Scott,

Thank you for your suggestions. In the meantime I was trying that in a different way. I have changed the prepare_amber.pl script in such a way that it will skip the ligands with bad geometries with just a warning (it won't stop there!!). Later I have removed those ligands from the input file for the Amber based run. Hope it will work.

I have reported earlier some problem with re screening using the same parameters. Few compounds got through with the first run could not make it after the second run (although I have used same parameters for both the runs, only difference was that the output of the first run was used as the input in case of second). Do you have any clue for that. I remember I had sent all the data files and parameter files to you earlier.

With best regards,
Rahul

-----Original Message-----
From: Scott Brozell [mailto:sbrozell at rci.rutgers.edu]
Sent: Tue 6/30/2009 2:22 PM
To: Rahul Banerjee
Subject: Re: [Dock-fans] prepare_amber.pl
 
Hi Rahul,

There has been some work on improving the sampling which will
be released in 6.4.  There was a fans where beta testing was suggested.
Look for the fans, maybe may, if u want to follow that path.

Scott

On Mon, Jun 01, 2009 at 05:20:21PM -0400, Rahul Banerjee wrote:
> Dear Scott,
> 
> I am having this problem when there is short contacts within the ligand. Earlier I have written to the Dock-fans about some strange geometries of ligands in the output. That same Dock output I am using for rescreening using Amber FF. I have attached one of such ligand coordinates I have got through Dock6. That same molecule gives problem with prepare_amber.pl due to its highly strained geometry (may be!!). Please find the input parameters in the Dock program. Is there any parameter in the Dock6 program to get rid of such bad geometries? 
> 
> Thanking you for your concern and suggestions.
> 
> Sincerely,
> Rahul
> 
> -----Original Message-----
> From: Scott Brozell [mailto:sbrozell at rci.rutgers.edu]
> Sent: Sun 5/31/2009 7:19 PM
> To: Rahul Banerjee
> Cc: dock-fans at docking.org
> Subject: Re: [Dock-fans] prepare_amber.pl
>  
> Hi,
> 
> On Fri, May 29, 2009 at 01:47:23PM -0400, Rahul Banerjee wrote:
> >  
> > I was trying to rescreen a set of ligand and rank them according to the Amber score. The script "prepare_amber.pl" worked nicely with the file provided in the Amber score tutorial. I was able to run it with another small test set. But it gives problem with other set. The program was stopped with the error message:
> >  
> > Ligand mayb.10 has total charge 0
> > Error from amberize_ligand; the name of the ligand is
> >     SMALL
> >     examine amberize_ligand.10.out
> > Error from amberize_ligand; examine amberize_ligand.10.out
> >  
> > If I check the file amberize_ligand.10.out it says:
> >  
> > Using ligand file prefix: mayb.10
> > Using antechamber options: -nc 0 -c bcc
> > Executing antechamber to generate the mol2 file with GAFF and charges.
> > 
> > Total number of electrons: 202; net charge: 0
> > Running: $AMBERHOME/bin/divcon
> > Error: unable to find divcon charges in divcon.out
> > Running: /opt/dock6/bin/bondtype -i ANTECHAMBER_BOND_TYPE.AC0 -o ANTECHAMBER_BOND_TYPE.AC -f ac -j full
> > Running: /opt/dock6/bin/atomtype -i ANTECHAMBER_AC.AC0 -o ANTECHAMBER_AC.AC -p gaff
> > Error running antechamber!
> >  
> > Can you please help me out? For other ligands (ligand.1 to ligand.9) this problem was not there. Is it because of any atom type that is not being recognized by the program?
> >  
> 
> There are several possible causes.
> Since AMBERHOME is defined and antechamber is spawning divcon,
> this faq is probably the issue:
> http://dock.compbio.ucsf.edu/DOCK_6/faq.htm#getpdb:%20can't%20open%20file%20lig.1.amber.pdb
> By undefining AMBERHOME, DOCK's mopac will be used.
> 
> However, if in addition there really are problems calculating charges then
> read this:
> http://blur.compbio.ucsf.edu/pipermail/dock-fans/2008-August/001700.html
> 
> Scott
> 



More information about the Dock-fans mailing list