[gmx-users] Bilayer explosion!

[David L. Bostick]

Hello all,

I have a system that consists of a 4 alamethecin helices in a DMPC bilayer
with water.  In order to build the system I took a neat DMPC Bilayer with
180 lipids which was well equilibrated and ran quite smoothly in GROMACS,
created a cavity within the bilayer and placed the 4 helices inside.  The 4
helices were obtained by stripping 4 helicies out of one of the starting
configurations Peter Tieleman supplies on his web site.  I then perform
many energy minimizations holding different groups still with constraints.
I then proceed to run a 200 ps simulation at 325K with all waters
restrained and the protein completely constrained.  This does not explode.
I then perform another 200 ps simulation at 325 K with only protein
constrained.  After about 150 ps in NPT the box starts expanding and the 2
bilayer leaflets and their water part ways leaving a bare, motionless
protein in the center of the simulation box.  In standard error I get 

Fatal error: ind = 13041 should be in 0 .. 12795 [FILE nsgrid.c, LINE 234]
Error on processor 0, will try to stop all the processors

I can only assume that it is because there is a big hole in the system and
gmx doesn't know how to handle the grid.  This is most likely a forcefield
error in the protein because the bilayer runs just fine.  It has happenned
before for a different protein, but I found that the error in that system
was that the protein had a nonstandard residue and that in addition to
changing the protein.itp file I had created, I also had to add the
nonstandard residue name to the aminoacids.dat file.  This seemed to make
things run smoothly.  

For this system, however, all residues of alamethecin are in the standard
aminoacids.dat file.  Also, the forcefield for alamethecin was created by
pdb2gmx.  Does anyone have some ideas as to what exactly might be causing
this? Also, in general, what role does the aminoacids.dat play in the
preprocessing of the system in grompp?

Thanks in advance,
David

-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
David Bostick					Office: 262 Venable Hall
Dept. of Physics and Astronomy			Phone:  (919)962-0165 
Program in Molecular and Cellular Biophysics 
UNC-Chapel Hill					
CB #3255 Phillips Hall				dbostick at physics.unc.edu	
Chapel Hill, NC 27599	           		http://www.unc.edu/~dbostick	
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-