[Dock-fans] dock6.6 and dock3.5 scoring

Trent E. Balius tbalius at aol.com
Fri May 30 10:14:18 PDT 2014


Hi Yunierkis,

With regards to electrostatics and charges.

That is weird.  I bet that this is something wrong with how you are 
preparing your receptor.
Perhaps, you can look at the test where the grids for dock3.5 score are 
prepared:

DOCK6_PATH/install/test/chemgrid_generation/

Look in the following for explanation of what commands are run for the 
tests.
  DOCK6_PATH/install/test/chemgrid_generation/Makefile
  DOCK6_PATH/install/test/default_targets.h


With regards to Amsol.

We use a modified version of amsol 6 in the Shoichet lab.

Just for your information, these modifications include the following:
(1) The ability to read in multiple ligands at a time.
(2) A method for dealing with z-matrix degeneracies.
(3) modifying the output to make parsing easier.

We are in the process of upgrading to the latest version Amsol 7.1 to 
make things easier for users.

There is a new version of DOCK 3.7 with some scripts for ligand 
preparation,
but these still use this modified Amsol 6 though.

You may be able to modify these scripts for your purposes.

I hope this helps,

Trent


---- Original Message ----
From: yunierkis <yunierkis at uclv.edu.cu>
To: William Joseph Allen <william.joseph.allen at gmail.com>
Cc: Trent E. Balius <tbalius at aol.com>; dock-fans <dock-fans at docking.org>
Sent: Fri, May 30, 2014 10:20 am
Subject: Re: [Dock-fans] dock6.6 and dock3.5 scoring


Hi Trent and Joe

Using Delphi5.1 fixed the issue with the electrostatic grid in the
examples provided with Dock6.6. However I'm still getting 0 for
electrostatic interactions.
Ligands come from a previous grid scoring dock6.6 run and have partial
charges. It is weird that I get no error during grid preparation and
conversion steps. Also I'm able to load and play with in Chimera the
generated Delphi generated electrostatic grid.

I have one additional question. Ligands atomic desolvation have to be
computed with AMSOL (I compiled AMSOL 7.1), I tried to use the scripts
provided in:

Wei BQ, Baase WA, Weaver LH, Matthews BW, Shoichet BK. A model binding
site
for testing scoring functions in molecular docking. J Mol Biol. 2002 
Sep 13;
322(2):339-55.

but the RunAMSOL script freezes when executing AMSOL, it just writes:

running AMSOL for water solvent

and nothing else happens.

Are there any other scripts that can be used to add atomic desolvation 
to the
ligands with AMSOL7.1?

Thanks
Yunierkis

On Thu, 2014-05-29 at 10:25 -0400, William Joseph Allen wrote:
> Hi Yunierkis and Trent -
>
>
> I thought I would jump in here for a quick comment. You may want to
> double-check that you remembered to add partial charges to both the
> ligand and receptor prior to generating grids / docking. Neglecting
> either of those will always result in null electrostatic energy
> scores.
>
>
> In the class we teach at Stony Brook, many of the students have this
> exact same problem with electrostatic energies. 99% of the time, they
> forgot to add partial charges to something.
>
>
> Joe Allen
>
>
> On Wed, May 28, 2014 at 10:05 AM, Trent E. Balius <tbalius at aol.com>
> wrote:
>         Hi Yunierkis,
>
>         It looks like from your error message that there is a problem
>         reading
>         in the
>         solvmap grid.  Was this generated correctly?
>
>         That is weird that you are get electrostatics energies of
>         zero.  This
>         implies that there is an issue with your electrostatic grids;
>         probably,
>         the error occurs when the delphi grid is converted
>         (grid-convrds, your
>         step 7).
>
>         the dock3.5 score is evaluated in the following test
>         directory:
>
>         $DOCK_PATH/install/test/chemgrid_score/
>
>         Are your electrostatic components  zero for all tests (dockout
>         files)?
>
>         Does chemgrid4.dockin run.  This test includes the ligand
>         desolvation?
>
>         Note that I was able to run these tests using DelPhi V. 5.1
>
>         I hope that this helps,
>
>         Trent
>
>         ======================================
>         Trent E. Balius, PhD
>         Shoichet Lab – Faculty of Pharmacy,
>         University of Toronto
>         Office Phone: (416) 946-7304
>         Cell Phone: (516) 381-0926
>         URL: http://docking.org/~tbalius
>         ======================================
>
>
>         ---- Original Message ----
>         From: yunierkis <yunierkis at uclv.edu.cu>
>         To: dock-fans <dock-fans at docking.org>
>         Sent: Fri, May 16, 2014 10:06 am
>         Subject: [Dock-fans] dock6.6 and dock3.5 scoring
>
>
>         Hi
>
>         I've been trying the dock3.5 scoring feature withing dock6.6
>         and I've
>         faced some problems. I'm following the examples provided with
>         the dock
>         6.6 installation. In brief, these are the steps I followed to
>         produce
>         the necessary grid files
>
>         1) Generate a receptor pdb file containing only polar Hs (with
>         chimera).
>         2) Rename all Hs to match the prot.table.ambcrg.ambH file
>         definitions.
>         3) Run solvmap
>         4) Run solvgrid. For both solvmap and solvgrid I used a grid
>         spacing of
>         0.25 A
>         5) Generate the selected spheres pdb file an merge with the
>         receptor
>         pdb file (rec+sph.crg)
>         6) Run delphi (v 6.2)
>         7) Run grid-convrds
>
>         Until here I got no error and all output files seems to be
>         correctly
>         created.
>
>         Then, I run dock6 with the following dock3.5 scoring specific
>         settings:
>
>         dock3.5_grd_prefix
>         ../../grid3.5_v1/chem52
>         dock3.5_electrostatic_score
>         yes
>         dock3.5_ligand_internal_energy
>         no
>         dock3.5_ligand_desolvation_score
>         volume
>         dock3.5_solvent_occlusion_file
>         ../../grid3.5_v1/solvmap
>         dock3.5_redistribute_positive_desolvation
>         no
>
>         And I get the following error:
>
>         Reading chemgrids from ../../grid3.5_v1/chem52.cmg
>         Reading DelPhi grid from ../../grid3.5_v1/chem52.phi
>         Reading ligand desolvation grid from ../../grid3.5_v1/solvmap
>
>
>         Memory bounds exceeded for solvent occlusion grid. Program
>         will
>         terminate.
>
>         The examples provided with dock 6.6 installation seems to work
>         fine but
>         I think I'm also having some problems with the delphi grids.
>         In the case of the examples I always get 0.0 for electrostatic
>         scores
>         which should not be correct.
>         I'm able to load the electrostatic potential map generated
>         with delphi
>         in chimera and associate it with my receptor. However when I
>         attempt to
>         load the chem52.phi file into chimera, I got the error:
>
>         Error reading file chem52.phi, format delphi
>         Negative record size -1053817856
>
>
>         Have anyone experienced these issues with the dock3.5 score?
>         How should
>         I proceed to make dock3.5 scoring work?
>
>         Thanks in advance for any help
>
>         Yunierkis
>
>
>
>         ------------------------------------------------------------
>         Universidad Central "Marta Abreu" de Las Villas.
>         Fundada el 30 de noviembre de 1952. Visítenos en:
>         http://www.uclv.edu.cu
>
>
>
>
>         _______________________________________________
>         Dock-fans mailing list
>         Dock-fans at docking.org
>         http://mailman.docking.org/mailman/listinfo/dock-fans
>
>
>
>         _______________________________________________
>         Dock-fans mailing list
>         Dock-fans at docking.org
>         http://mailman.docking.org/mailman/listinfo/dock-fans
>
>
>
>
> --
> William Joseph Allen, Ph.D.
> Postdoctoral Associate
> Dept. of Applied Mathematics and Statistics
> Stony Brook University
> http://ringo.ams.sunysb.edu/~wjallen/
>
>
>
> ______________________________________________________________________
> Universidad Central "Marta Abreu" de Las Villas.
> Fundada el 30 de noviembre de 1952. Visítenos en:
> http://www.uclv.edu.cu
>
>



Universidad Central "Marta Abreu" de Las Villas.
Fundada el 30 de noviembre de 1952. Visítenos en:  
http://www.uclv.edu.cu



  


More information about the Dock-fans mailing list