[gmx-users] mdrun -rerun and box size bug not fixed?

[Michel Cuendet]

Hi gromacs people,

I was trying to rerun old xtc trajectories to recalculate some energies 
involving the solvent, using gmx 3.3.1. I saw important discrepancies 
with the original energies (nstlist=1, xtc unaccuracy tolerated). After 
some struggling, I figured out that the box size is determined form the 
.gro file used to create the .tpr file, and are not read from the 
trajectory. I then realized that this problem had been described two 
years ago on the list:

    http://www.gromacs.org/pipermail/gmx-users/2005-October/017729.html

The issue was then transferred to bugzilla :

    http://bugzilla.gromacs.org/show_bug.cgi?id=26

and a patched md.c was proposed by David, apparently for gromacs 3.3.0. 
It seems that this patch has not been integrated in gmx 3.3.1, because 
the -rerun option still gives wrong energies. What's worse, the patched 
md.c cannot be used for 3.3.1, because apparently the number of 
arguments for init_md() has changed.

Could someone work out a patch for 3.3.1, and maybe post it on the same 
bugzilla thread?

I think this bug should be mentioned in the "known bugs" section of the 
website. That would probably save some sweat to others like me trying to 
use the -rerun option...

Thanks,
Michel

-- 

==========================================================
Michel Cuendet, Ph.D
Molecular Modeling Group
Swiss Institute of Bioinformatics
CH-1015 Lausanne, Switzerland
www.isb-sib.ch/groups/Molecular_Modeling.htm
==========================================================