LAMMPS_INSTRUCTION.txt

Prepared by Weihua Zheng

******************
A. Install LAMMPS:
1. download LAMMPS: http://lammps.sandia.gov/download.html (the tarball version, not the SVN version)
2. download fftw-2.1.5. and install it. http://www.fftw.org/download.html (newer version might have problems)
	./configure --prefix=/home/weihua/fftw-2.1.5
	make
	make install

3. cd LAMMPS/src/STUBS/ (LAMMPS/ is where the tar ball is extracted to)
   make

4. copy the attached Makefile.serial to LAMMPS/src/MAKE/, then 
LAMMPS/src/make serial ('make clean-all' before recompile)
'make serial' if Makefile.serial is used.

**************************************
B. Prepare input for LAMMPS simulation
1. Install BioPython: http://biopython.org/wiki/Download
2. Get a copy of the LAMMPS_tools developed by Aram (attached).
3. Make a new directory 'newdir/', copy all scripts in LAMMPS_tools/create_project_tools/ into the directory. Copy all files in LAMMPS_tools/parameters/ to 'newdir/'.
4. Copy the pdb file you want to study into the directory, run 
./Pdb2Lammps.sh PDB_ID input (Note: PDB_ID is the pdb file name without '.pdb' part)
The code will take the pdb file and generate a couple of input files for the simulation.
5. Currently, 'ssweight' input file needs to be generated manually. Here is how. 
 Submit your sequence to 'Jpred3' at http://www.compbio.dundee.ac.uk/www-jpred/. Use the 'view simple' results to generate the file. 'ssweight' is a two column data, number of rows is the same as the number of residues. In each row, if the corresponding residue is in helix form ('H'), then write '1.0 0.0'; write '0.0 1.0' if it's in beta sheet.
6. Edit input.in file, change 'sequance.se' to 'input.se' or whatever input name you used in step 4. Find 'nvt' and add 'temp' after it with a space in between. After line 'dump ...', add a new line 'dump_modify 1 sort id' so that the atom order is correctly written into the trajectory file.

*************************
C. Run LAMMPS simulation
In 'newdir/' created above, run
LAMMPS/src/lmp_serial < input.in
'lmp_serial' is the exe compiled in step A.