[Chimera-users] [Dock-fans] Help with antechamber

[Scott Brozell]

Hi,

On Tue, Mar 06, 2012 at 08:04:51PM -0500, Brad Ridder wrote:
> Thanks for the responses, but I am still stuck on antechamber.
> 
> We are holding off on this python scripting; that was the plan before, but
> I want to make sure I can get the individual components working properly
> before I go to the effort of writing a python script to do dock prepping.
> 
> I cannot get either the DOCK-supplied version of antechamber or the
> CHIMERA-supplied version of antechamber to work correctly.

==== dock
> First, I tried the DOCK distribution again. $AMBERHOME was not defined. I
> used:
> 
> *$DOCK_HOME/bin/antechamber -i $PDBFILES/3GLR.pdb -fi pdb -o
> $DESKTOP/3GLRcharged.mol2 -fo mol2 -c bcc*
> 
> This returns the error:
> *
> Total number of electrons: 0; net charge: 0
> 
> Running: $DOCK_HOME/bin/mopac.sh
> Error: unable to find mopac charges in mopac.out
> *

It is possible that antechamber is working and that mopac really fails.
Did you look in mopac.out ?
Are you really trying to run antechamber on 3glr, which is a protein ?
antechamber is not the way to assign charges to a protein.
Read the AmberTools manual
http://ambermd.org/doc11/AmberTools.pdf

Here is one way to test your dock's antechamber:
cd dock6/install/test/amber_score_181l  
make
../../../bin/antechamber -s 2 -i lig.1.mol2 -fi mol2 -o lig.1.gaff.mol2 -fo mol2 -nc 0 -c bcc
This last command should emit:
...
Running: $DOCK_HOME/bin/mopac.sh

Running: .../dock6/bin/am1bcc -i ...
...


> I then repeated this later with $AMBERHOME defined to $CHIMERA/bin/amber11,
> which gives the new error message:
> 
> *Total number of electrons: 0; net charge: 0
> 
> Running: $AMBERHOME/bin/divcon
> sh:
> /home/abhay/Desktop/rcg-research/personal-work/CHIMERA/bin/amber11/bin/divcon:
> not found
> Error: cannot run "$AMBERHOME/bin/divcon" of bcc() in charge.c properly,
> exit*

This approach - using dock's antechamber with amberhome defined
to be inside chimera - is not a good one.
However, you might get it to work:
It is very likely that sqm, not divcon, is the qm program that should
be invoked.  So get help to look for the option to force sqm usage:
antechamber -h
...
                   -df  am1-bcc flag, 2 - use sqm(default); 0 - use mopac
...
is probably what you'll find; if so then add -df 2 to your antechamber 
command line.

> Now, a few questions/reflections:
> 
> 1. As per the first error message, why is mopac.out causing trouble? Is
> this a file I am supposed to make?

mopac.out is the output of the mopac qm program which antechamber
invokes and then processes to make am1-bcc charges.  

> 2. As per the second error message, I have no idea what "divcon" is, or
> what it is needed for. There is no divcon file that comes with the CHIMERA
> amber11 directory, nor is there any divcon file in the AmberTools amber11
> directory I have on my Desktop.

divcon is another qm program.


==== chimera
> I then tried the version of antechamber with the CHIMERA distribution.
> First I typed:
> 
> *$CHIMERA/bin/amber11/bin/antechamber*
> 
> And antechamber said "AMBERHOME is not set!"
> 
> So, I set $AMBERHOME to be $CHIMERA/bin/amber11. Then I used
> 
> *$CHIMERA/bin/amber11/bin/antechamber -i $PDBFILES/3GLR.pdb -fi pdb -o
> $DESKTOP/3GLRcharged.mol2 -fo mol2 -c bcc*
> 
> The Terminal then spams:
> 
> *Warning: detected more than 10 Residue sequence numbers;
>     this may be a large multiple residue PDB file;
>     large multiple residue PDB files are not supported.
>     Continuing, but problems may be encountered.*
> 
> And then finally stops with
> 
> *Segmentation fault*

This warning was absent in early versions of antechamber.
And it likely right on target.  mopac (above) probably failed 
because you are really trying to run antechamber on 3glr.
antechamber is not the way to assign charges to a protein.

> So, in short, I am utterly bamboozled. Obviously, antechamber can handle
> "large multiple residue PDB files", since the Dock Prep tutorial used the
> AM1-BCC charge model on the large 1ABE.pdb protein, and it worked fine. Why
> is it not working for me?

No, the receptor charges were assigned using residue matching.
See step 1 of
http://dock.compbio.ucsf.edu/DOCK_6/tutorials/struct_prep/prepping_molecules.htm
and the chimera dockprep docs:
http://www.cgl.ucsf.edu/chimera/current/docs/UsersGuide/framecontrib.html
"Charges for standard residues ... are taken from Amber (details)."
Read the force field chapter of the AmberTools manual and references
therein for even more details.


scott

> On Wed, Feb 29, 2012 at 10:30 PM, Scott Brozell <sbrozell at rci.rutgers.edu>wrote:
> > Here's another similar ligand prep script:
> > http://mailman.docking.org/pipermail/dock-fans/2011-December/002591.html
> > While it is a good pedagogical exercise to prep once manually,
> > scripting is the only practical approach for lots of ligands.
> >
> > As far as running antechamber, be careful and be specific about which
> > one you run:
> > If you are manually running Dock's antechamber then define DOCK_HOME.
> > Do not define AMBERHOME.
> > See
> > http://dock.compbio.ucsf.edu/DOCK_6/faq.htm#Error%20from%20amberize_ligand
> > ;
> >
> > If you are manually running Chimera's antechamber then DOCK_HOME
> > is irrelevant but AMBERHOME may be relevant.
> > ( I have not manually run Chimera's antechamber, so my advice may be
> > wrong. )
> >
> > Also note that Dock's antechamber is circa 2006 with some key subsequent
> > bugfixes.  Chimera's antechamber is much more recent.
> > Dock's antechamber uses mopac for qm;
> > Chimera's antechamber uses sqm for qm;
> > (http://ambermd.org/#AmberTools)
> > I have found mopac to be more robust than sqm.
> > Your mileage may vary.
> >
> > scott
> >
> > On Wed, Feb 29, 2012 at 04:08:22PM -0800, Eric Pettersen wrote:
> > > This previous posting to the dock-fans list seems to be more directly
> > > pertinent:
> > >
> > > http://mailman.docking.org/pipermail/dock-fans/2007-May/001043.html
> > >
> > > On Feb 29, 2012, at 2:47 PM, Trent E. Balius wrote:
> > > >I am cc'ing this to the chimera list.
> > > >
> > > >I think that you should write a python script to run DOCK prep on
> > > >the command line.  See the following:
> > > >
> > > >http://www.cgl.ucsf.edu/pipermail/chimera-dev/2011/000784.html
> > > >
> > > >For other example scripts see the following:
> > > >
> > > >
> > http://www.cgl.ucsf.edu/pipermail/chimera-users/2009-February/003539.html
> > > >
> > > >You should not have to specify the AMBERHOME path.  This is done
> > > >internally to chimera.
> > > >
> > > >as an aside, It is unnecessary to run antichamber with sudo.
> > > >
> > > >-----Original Message-----
> > > >From: Brad Ridder <clustro at gmail.com>
> > > >To: dock-fans <dock-fans at docking.org>
> > > >
> > > >My name is Brad; I am a grad student in the dept. of chemical
> > > >engineering at Purdue University.
> > > >
> > > >We are experimenting with DOCK for various drug discovery work. At
> > > >the moment, I am trying to do execute all of the "dock prep"
> > > >commands through the Terminal, since it is not practical to do large
> > > >volumes of Dock Prep by hand through the CHIMERA GUI.
> > > >
> > > >Corresponding to what is in the DOCK tutorial, I am on "Receptor and
> > > >Ligand Preparation." I got reduce to work, in order to add hydrogens.
> > > >
> > > >Now, I am trying to get antechamber to work, in order to add the
> > > >charges, but it is not working. Here is the command I am running:
> > > >
> > > >sudo ./antechamber -i $DESKTOP/3GLR.pdb -fi pdb -o $DESKTOP/
> > > >3GLRcharged.mol2 -fo mol2
> > > >
> > > >Which returns the error:
> > > >
> > > >Warning: $AMBERHOME and $DOCK_HOME enviornment strings are not set,
> > > >use "mopac.sh" in the work directory
> > > >
> > > >Cannot open CONNECT.TPL , exit
> > > >The atom number exceeds the MAXATOM, reallocate memory
> > > >
> > > >I looked around in the dock6 directory, and found a folder called
> > > >"parameters/antechamber", which contains CONNECT.TPL. I figured this
> > > >is what should be the $AMBERHOME directory.
> > > >For DOCK_HOME, I figured it should be the dock6 folder.
> > > >
> > > >Here is what I have in my .bash_profile file in my home folder.
> > > >
> > > >export AMBERHOME='/home/.../DOCK6.5/dock6/parameters/antechamber'
> > > >export DOCK_HOME='/home/.../DOCK6.5/dock6'
> > > >
> > > >I also tried using the "amber11" folder that comes with CHIMERA as
> > > >AMBERHOME, but to no effect.
> > > >
> > > >I don't actually have AMBER installed on my computer, but since
> > > >CHIMERA can add charges without actually requiring the purchase of
> > > >AMBER, I do not think that is the problem.
> > > >
> > > >I did not know what to do with "mopac.sh." I ran the script, but it
> > > >says it needs input files.
> > > >
> > > >What am I doing wrong that is preventing antechamber from executing
> > > >correctly?