Hydrophobic Gating with Gromacs 3
Reference: Oliver Beckstein, Philip C. Biggin, and
Mark S. P. Sansom; A Hydrophobic Gating Mechanism for
Nanopores, The Journal of Physical Chemistry B; 2001;
105(51); 12902-12905.
[abstract]
OVERVIEW:
0 Unpack input files
1 Locations and Directory-naming convention
1.1 general
1.2 System specific
2 Files to be modified
3 Running a simulation
I generated input files for different pore radii from 3.5 to 10
A. This includes the pdb for the solvated system.
I am using (GNU) Makefiles to do most of the work so I also included
the appropriate Makefiles. I describe what to change below (section 2)
so you might be able to use them as well. In any case, you can have a
look at them to find out how to run things manually.
All input files are in the tar file hygate.tar.bz2 (or hygate.tar.gz).
0 Unpack input files
Download tar archive from
http://indigo1.biop.ox.ac.uk/users/oliver/download/HyGate/hygate.tar.bz2
Unpack with
tar -jxvf hygate.tar.bz2
cd HyGate
(or tar -zxvf hygate.tar.gz)
1 Locations and Directory-naming convention
1.1 General
You need mth.itp (defines the CH4 pseudoatoms in the pore) and lay.itp
(the pseudoatoms in the membrane mimetic slab). Put them in a
directory of your choice, e.g. the gromacs topology directory. Or put
it somewhere else and add the directory to the include list of the mdp
files (see 1.2).
killwater.pl is a script which removes waters between z1 and z2; it is
needed to generate input for systems with an initially empty
cavity. If required it should be in the shell's search path.
vdwradii.dat is the standard van der Waals radii data file plus CH4
(r=0.195nm). It is used for the generation of the pdb files and is
referenced as VDWDAT_SRC in the Makefile. You probably don't need it.
1.2 System specific
The setup files live in directories whose pathname indicates the
parameters of the system as for instance pore radius, pore length etc.
The generic directory names is
G<n>/L_<l>/<r>/<q>/<c>/<identifier>
<...> stands for a parameter.
parameters:
Geometry n (R_outer, R_mouth, L_mouth = 4A)
1 = {15.5, 7.5},
2 = {18, 10}, 2a = {19.5, 10},
3 = {26, 22}
4 = {25.8, 18}
Length l (L_pore): 8 A, 12 A
Radius r (R_pore): 3.5 ... 14 A
Charge q
0 hydrophobic (no charge)
D1 one dipole (parallel z)
D2 two dipoles (both parallel z: down/down)
D2a two dipoles (antiparallel z: up/down)
D4 four dipoles (all parallel z)
Q2 two quadrupoles
Cavity c: start with water filled (full) or empty (empty)
pore cavity
Identifier
a unique name; I used it for runs starting with
different random seeds in the initial velocity
distribution.
2 Files to be modified
You have to change:
* mdp files: chl_emin.mdp
chl_md.mdp
'include' has to point to the directory which contains mth.itp and
lay.itp unless they are already in Gromacs search path
* Makefile.inc
NPROC: number of processors to generate tpr files for
Optional:
* Makefile
Makefile contains a few other paths that are specific to my computer
but you'll probably not need it anyway. You would need it (plus some
other programs and scripts) if you wanted to generate input files
and pdbs from scratch.
* rungmx.bash
If you actually want to submit the jobs to a queuing system (we are
using PBS here) you will have to edit a fair number of paths and the
'#PBS -l' options for the number of nodes/processors used in
rungmx.bash.
If everything is set up properly it is simply 'qsub rungmx.bash'.
3 Running a simulation
If everything is set up up properly you should be able to generate
an input tpr file for the energy minimisation with
make emin
Do the energy minimisation
make run_emin
Generate an input tpr file for the MD
make md
and start the simulation
mpirun C -v -c2c -np 2 mdrun -np 2 -v \
-s chl_md.tpr \
-deffnm chl_lay_water_md
Please refer to the Makefile for the explicit syntax of the first
three steps.
You can also look through rungmx.bash: In the last quarter of the file
(after lots of house keeping tasks) the same steps are executed.
Last modified: 2002-01-07 15:27:19 GMT
by Oliver Beckstein