1
# Use these commands to generate the LAMMPS input script and data file
2
# (and other auxilliary files):
4
# First, generate the coordinates (the "system.xyz" file) this way
5
# (This can be a very slow process.)
8
# You must run each packmol commend one after the other
9
# NOTE: If PACKMOL gets stuck in an endless loop, then edit the corresponding
10
# "inp" file. This should not happen. You can also usually interrupt
11
# packmol after 30 minutes, and the solution at that point should be good
13
packmol < step1_proteins.inp # This step determines the protein's location
14
# It takes ~20 minutes (on an intel i7)
15
packmol < step2_innerlayer.inp # this step builds the inner monolayer
16
# It takes ~90 minutes
17
packmol < step3_outerlayer.inp # this step builds the outer monolayer
20
# Running these three commands will probably take about an hour.
21
# (It depends on how uniformly you need the packing to be.)
23
mv -f step3_outerlayer.xyz ../system.xyz
27
# Create LAMMPS input files this way:
32
moltemplate.sh -xyz ../system.xyz system.lt
34
# This will generate various files with names ending in *.in* and *.data.
35
# These files are the input files directly read by LAMMPS. Move them to
36
# the parent directory (or wherever you plan to run the simulation).
38
mv -f system.in* system.data ../
40
# The "table_int.dat" file contains tabular data for the lipid INT-INT atom
41
# 1/r^2 interaction. We need it too. (This slows down the simulation by x2,
42
# so I might look for a way to get rid of it later.)
43
cp -f table_int.dat ../
46
# The "./output_ttree/" directory is full of temporary files generated by
47
# moltemplate. They can be useful for debugging, but are usually thrown away.