[Dock-fans] grid-convert segmentation fault + SOLVMAP out of range

Collignon, Barbara C. collignonbc at ornl.gov
Fri May 1 08:36:06 PDT 2009


Hi Scott,

Thanks again for having resolved the grid-convert problem. 

So, as I said before, I still get this problem with solvmap

Do you think it might be because of the size of the grid ?
If yes, where should I change in the code the size of the 
table to be allocated ?

Initializing Library File Routines...
Initializing Orienting Routines...
Initializing Conformer Generator Routines...
Initializing Grid Score Routines...
 Reading the energy grid from grid.nrg
Initializing Dock3.5 Score Routines...
 Reading chemgrids from chem52.cmg
 Reading DelPhi grid from chem52.phi
 Reading ligand desolvation grid from solvmap
terminate called after throwing an instance of 'std::out_of_range'
  what():  basic_string::substr
Abort

Barbara C. Collignon, Ph.D
Post-doctoral Research Fellow
Center For Molecular Biophysics
UT-Oak Ridge National Laboratory
Building 6011, Mail Stop 6309
Oak Ridge, TN  37831-6309

Phone:865-241-5175

https://cmb.ornl.gov/group/bci



-----Original Message-----
From: Scott Brozell [mailto:sbrozell at rci.rutgers.edu]
Sent: Thu 4/30/2009 9:08 PM
To: Collignon, Barbara C.

Barbara C. Collignon, Ph.D
Post-doctoral Research Fellow
Center For Molecular Biophysics
UT-Oak Ridge National Laboratory
Building 6011, Mail Stop 6309
Oak Ridge, TN  37831-6309

Tel:865-241-5175

https://cmb.ornl.gov/group/bci



-----Original Message-----
From: Scott Brozell [mailto:sbrozell at rci.rutgers.edu]
Sent: Thu 4/30/2009 9:08 PM
To: Collignon, Barbara C.
Cc: dock-fans at docking.org
Subject: Re: [Dock-fans] grid-convert segmentation fault
 
Hi,

The seg fault only occurs when
contact_score                  yes
because grid-convert logically should never convert a contact grid
since that grid is already in DOCK readable form.
The programmer(s) that modified grid to produce grid-convert removed
the contact grid reading code but did not remove the contact grid 
writing code nor correct the user interface nor document its behavior.

Ive improved the behavior and docs, but for 6.2 the fix is to
workaround this by inputing
contact_score                  no

thanks for the report,
Scott

On Wed, Apr 29, 2009 at 01:46:49PM -0400, Collignon, Barbara C. wrote:
> 
> I see...The error that you get should be related to the fact that DELPHI is compiled in 32 bit while DOCK6.2 is in 64 bit.
> 
> What I did:
> 
> On my laptop (32 bit platform), I compiled DOCK6.2 with gnu and I use the binary code of DELPHI for 32 bit.
> Linux-5qvm : /home/barbara  > uname -a
> Linux linux-5qvm 2.6.25.18-0.2-default #1 SMP 2008-10-21 16:30:26 +0200 i686 i686 i386 GNU/Linux
> 
> On the linux cluster (64 bit platform), it is working ONLY IF DOCK6.2 is compiled in 32 bit. DELPHI remains
> in 32 bit.
> bioinfo : /data1/bci/DOCK_desolv/DOCK_ace/RIGID/TEST_3.5  > uname -a
> Linux bioinfo.lsd.ornl.gov 2.6.9-42.0.2.ELsmp #1 SMP Wed Aug 23 13:38:27 BST 2006 x86_64 x86_64 x86_64 GNU/Linux
> 
> -----Original Message-----
> From: Scott Brozell [mailto:sbrozell at rci.rutgers.edu]
> 
> Unfortunately, I cannot even get as far as you using DelPhi II on x86_64;
> What delphi and platform are you using ?
> 
> % grid-convert -i gconv.in
> ...
> Reading in grid box information.
> Box center of mass                      :   41.483   33.072   48.120
> Box dimensions                          :   34.187   37.462   37.293
> Number of grid points per side [x y z]  :      115      126      126
> Total number of grid points             :  1825740
> 
> Reading scoring grids.
>  1339728  0.333333313 104 114 113  24.3895016  14.3409977  29.4734993
>  maxgrid  2500000
> do_us: off end of record
> apparent state: unit 2 named rec+sph.phi
> last format: list io
> lately reading sequential unformatted external IO
> Abort
> 
> % uname -a
> Linux edu 2.6.18-92.1.22.el5 #1 SMP Tue Dec 16 11:57:43 EST 2008 x86_64 x86_64 x86_64 GNU/Linux
> $ ~/dock/help  delphi
> 
>   ___________________DelPhi II____________________
>  /
>  | A program to solve the PB equation             |
>  | in 3D, using non-linear form, incorporating    |
>  | 2 dielectric regions, ionic strength, periodic |
>  | and focussing boundary conditions, utilizing   |
>  | stripped optimum successive over-relaxation    |
>  | and an improved algorithm for mapping the      |
>  | Mol. Surface to the finite-Difference grid     |
>  __________________          ____________________/
>                      DelPhi II
> 
>   program started on Tue Apr 28 2009
>               at 19:33:55
>  parameter file fort.10 does not seem to     exist
>  nonexistence or unexpected end of radius file
> FORTRAN STOP
> 
> On Tue, Apr 21, 2009 at 07:53:19PM -0400, Collignon, Barbara C. wrote:
> > 
> > I did not find any solutions:
> > 
> > I still get:
> > 
> > Reading scoring grids.
> >  1339728  0.333333313 104 114 113  24.3895016  14.3409977  29.4734993
> >  maxgrid  2500000
> > Writing general grid info to chem52.bmp
> > Writing bump grid to chem52.bmp
> > Writing Delphi grid info to chem52.phi
> > Writing Delphi grid to chem52.phi
> > Writing contact grid to chem52.cnt
> > Segmentation fault
> > 
> > I sent you my input files, if you want to reproduce the bug...
> > But maybe it is just a compilation problem.
> > 
> > I used to execute ./script_grid to build the pre-conputed grids
> > 
> > Barbara
> > 
> > 
> > -----Original Message-----
> > From: Scott Brozell [mailto:sbrozell at rci.rutgers.edu]
> > Sent: Mon 4/20/2009 9:19 PM
> > To: Collignon, Barbara C.
> > Cc: dock-fans at docking.org
> > Subject: Re: [Dock-fans] grid-convert segmentation fault
> >  
> > Hi,
> > 
> > On Fri, 17 Apr 2009, Collignon, Barbara C. wrote:
> > 
> > > I am using DOCK6.2, the compiled version for 32bit,
> > > on both my laptop and our local Linux cluster.
> > >
> > > after generating the delphi grid, solvmap, chemgrids, contact
> > > and energy grids, I do >grid-convert -i gconv.in
> > >
> > > At this point, I do not get errors anymore when reading the delphi grid.
> > > Instead, I get a "segmentation fault", when writing the contact grid
> > > to chem.cnt (from grid.cnt). (the same error on both my laptop and
> > > our local Linux cluster)
> > 
> > Did you fix this issue ?
> > What was the problem ?
> > 
> > thanks,
> > Scott



More information about the Dock-fans mailing list