[gmx-users] Weird structure after minimization (membrane proteinsimulation

[Mark Abraham]

alokjain at iitk.ac.in wrote:
> Dear Mark,
> 
> Thanks a lot for your reply.
> 
> Default was 1.6 angstrom and when I change gradually till 2.5 angstrom then
> structure looks similar to original structure.
> 
> I have again checked in my top and mdp files for include statement.
> 
> I have define = -DFLEXIBLE in my mdp file

Good, that's a material difference that you can use to find where the 
problem occurs, since this is only present in your EM, not your simulation.

> in my top file I have included the tip4p.itp file as follow.
> #include "tip4p.itp"
> 
> which must open the tip4p.itp file comes with the distribution which has
> following section

This bears little resemblance to a tip4p.itp file I have in a 3.3.1 
distribution I had lying around.

> #######################################################################
> 
> [ moleculetype ]
> ; molname nrexcl
> SOL  2
> 
> [ atoms ]
> ; id at type res nr  residu name at name cg nr charge
> #ifdef _FF_OPLS
> 1 opls_115 1 SOL MW1 1      -1.04
> 2 opls_114 1 SOL HW2 1        0.52
> 3 opls_114 1 SOL HW3 1        0.52
> 4 opls_113 1 SOL OW4 1        0

There should be more #ifdef regions here for various force fields, and 
the atom order is changed from my file. I think this might introduce a 
problem if the atom order in your water molecules in your structure file 
differs, but I'm guessing here.

> #ifdef FLEXIBLE
> [ bonds ]
> ; i j funct length force.c.
> 4 2 1 0.09572 502416.0 0.09572 502416.0
> 4 3 1 0.09572 502416.0 0.09572 502416.0

These are well-formed, given the above, but if you've given the whole 
tip4p.itp file then there are missing #endifs such that it isn't working 
properly.

> [ angles ]
> ; i j k funct angle force.c.
> 2 4 3 1 104.52 628.02 104.52 628.02

Likewise.

> ########################################################

I'd start by finding out where you got this tip4p.itp file, and probably 
hitting someone over the back of the head for messing around 
unnecessarily with it. Or, if it was necessary to change it for some 
purpose, hitting them over the back of the head for letting you have a 
copy of this file, not a normal one :-)

Then, I'd get a clean distribution, start from the beginning, check the 
tip4p.itp file and see how it goes.

Mark