[gmx-users] Minimization changes system box shape after using ParmEd

[Crystal Vander Zanden]

Hello GROMACS community,

I am having problems maintaining the shape of my system box after using
ParmEd to convert from Amber to Gromacs format.

I parameterized a system with AmberTools (peptide solvated in an octahedral
box).  I then converted the Amber files (.inpcrd and .prmtop) to GROMACS
format (.gro and .top) using ParmEd.  I used VMD to look at the system, and
it looks ok (peptide still centered in an octahedral box). (
http://i1380.photobucket.com/albums/ah185/cmvander/parmed_zpssaq6ldnb.png)

Next I did an energy minimization in GROMACS.  VMD displayed the resulting
.gro file as a triclinic box and the peptide is no longer centered. (
http://i1380.photobucket.com/albums/ah185/cmvander/minimized_zpsekznpl3w.png
)

I think that the system really is still an octahedral box because I can see
the octahedral box boundaries within the triclinic box. I overlaid both
systems and the protein is in the same spot (
http://i1380.photobucket.com/albums/ah185/cmvander/image_overlay_zpsepfpqkvl.png).
Also I got reasonable minimization energies.

Steepest Descents converged to Fmax < 1000 in 337 steps
Potential Energy  = -3.1393012e+05
Maximum force     =  9.0675385e+02 on atom 363
Norm of force     =  3.5674736e+01

Am I just doing something wrong to cause VMD to display my system
incorrectly, or is there a more serious issue?  Either way, I’d be happy to
know how to fix it!

Thanks in advance for any advice!  I really appreciate it!
Crystal

Some extra information-------
The last line of the AMBER .inpcrd file:
68.4778914  68.4778914  68.4778914 109.4712190 109.4712190 109.4712190

The last line of the .gro file from ParmEd conversion:
6.84779   6.45616   5.59120   0.00000   0.00000  -2.28260   0.00000
-2.28260  -3.22808

The last line of the em.gro file resulting from energy minimization:
6.84779   6.45616   5.59120   0.00000   0.00000  -2.28260   0.00000
-2.28260  -3.22808

I tried using editconf as suggested here (
http://manual.gromacs.org/programs/gmx-editconf.html), but I don’t think
this is the right track for my problem. Minimization failed when I tried to
use the structure I made by doing this.
“To convert a truncated octrahedron file produced by a package which uses a
cubic box with the corners cut off (such as GROMOS), use:
*>>>*gmx editconf -f in -rotate 0 45 35.264 -bt o -box veclen -o out
where veclen is the size of the cubic box times sqrt(3)/2.”


-- 
Crystal M. Vander Zanden, Ph.D.
ASERT Postdoctoral Fellow (IRACDA)
University of New Mexico, Albuquerque