[gmx-users] genbox solvent addition inhomogeneous?

[Pim Frederix]

Dear users,

When I'm trying to create my solve box of 12x12x12 nm I encounter some trouble. I'm using Gromacs 4.5 with the MARTINI force field to set up a system of 400 dipeptide molecules in water. However when I use "genbox -cp <coarse grained dipeptidebox> -cs <MARTINI example water box> -vdwd 0.2 -o <output>" I get an inhomogeneous distribution of the water beads: there is what looks like a cubic higher density water box in the middle ~9 nm of the output. When I then try to run a energy minimization with mdrun the simulation crashes quickly with

Steepest Descents converged to machine precision in 15 steps,
but did not reach the requested Fmax < 100.
Potential Energy  =  7.8217890e+17
Maximum force     =            inf on atom 6750
Norm of force     =            inf

I've tried different vdwd's, using define = -DPOSRES, but I can't get round it. Using smaller 9x9x9 boxes with 150 dipeptides and further nothing different does work very satisfactory, but to study the assembled structure I would like to increase the size of the entire system.

Can anyone explain why this happens or suggest any solutions? Thanks very much,

Pim Frederix
PhD student University of Strathclyde
Pure & Applied Chemistry / Biomolecular & Chemical Physics