LAMMPS Simulation for Water using SPC/E Forcefield

In this tutorial we learn how to run an MD simulation for water and ice Ih.

Install GenIce2

pip3 install genice2

Install VMD

https://www.ks.uiuc.edu/Research/vmd/

Generate ice Ih configuration using GenIce2

genice2 1h -w physical_water --rep 2 2 2 --format gromacs > ice1h128.gro

Generate LAMMPS data file with VMD

Download the .gro file with scp:

realpath ice1h128.gro
scp <YourNetID>@adroit-vis.princeton.edu:PATH_TO_GRO .

Load the .gro file in VMD:

Let's take a look at topo.tcl by running cat topo.tcl:

set nH2O 128
set lO {}
set lH1 {}
set lH2 {}
for {set x 0} {$x <$nH2O} {incr x} {lappend lO [ expr 3*$x+0]}
for {set x 0} {$x <$nH2O} {incr x} {lappend lH1 [ expr 3*$x+1]}
for {set x 0} {$x <$nH2O} {incr x} {lappend lH2 [ expr 3*$x+2]}
puts $lO
puts $lH1
puts $lH2
set selO [atomselect top "index $lO"]
set selH1 [atomselect top "index $lH1"]
set selH2 [atomselect top "index $lH2"]
for {set x 0} {$x <$nH2O} {incr x} {
set id0 [lindex $lO $x];
set id1 [lindex $lH1 $x];
set id2 [lindex $lH2 $x];
topo addbond $id0 $id1;
topo addbond $id0 $id2;
topo addangle $id1 $id0 $id2;
}
$selO set charge -0.8476
$selH1 set charge 0.4238
$selH2 set charge 0.4238
topo writelammpsdata water.data

We may directly generate the data file in VMD's Tk Console by using

source topo.tcl

Upload the data file onto the cluster:

scp water.data <YourNetID>@adroit-vis.princeton.edu:PATH_TO_WATER_TUTORIAL

Run the simulation

The initial configuration of our simulation is ice Ih. At 300 K, you will observe ice Ih melting. Ice will remain to be ice at 100 K.

Now let's enter the working directory:

cd 300K

If you run MD using slurm, run the following command:

sbatch run.slurm

If not, just run LAMMPS:

lmp -in spec.lmp

Things are similar for 100 K.

If you are interested in exploring Deep Potential, a flavor of Machine Learning Forcefield, you may check:

https://github.com/CSIprinceton/workshop-july-2022/