Home - adavtyan/awsemmd GitHub Wiki

Getting started

Building LAMMPS with AWSEM

What do you need

Step by step guide

  • Download and unzip LAMMPS source code
  • If not already included, add the MOLECULE package, by executing make yes-molecule
  • Copy all .cpp and .h file from awsemmd directory to LAMMPS_DIR/src/
  • Go back to LAMMPS_DIR/src/ directory and execute make clean-all; make serial
  • If everything goes well, you must find lmp_serial executable in src directory. See the list of known issues here.

For more information on LAMMPS compilation and inclusion of optional packages see the following link http://lammps.sandia.gov/doc/Section_start.html#start_2

Generating project

The easiest way to generate a LAMMPS-AWSEM project is to do that using a pdb file which contains your structure of interest. You will need to have python and biopython installed on your computer.

  • First of all go to awsemmd/tools/create_project_tools/ and execute install_tools.sh script. This will copy all necessary scripts in /your_home_directory/opt/script/.
  • Add /your_home_directory/opt/script/ to PATH variable (call echo $PATH to make sure that you did it right)
  • Copy lmp_serial and the pdb file to a same directory
  • From that directory call PdbCoords2Lammps.sh pdb_id project_name

After this you should find the following files in your simulation (current) directory

  1. data. file which contains atom and bond descriptions
  2. .seq file with protein sequence
  3. .coord with all atom backbone description and which is not needed for LAMMPS-AWSEM simulations
  4. .in which contains an input script

You can modify the .in according your needs. To learn more about the commands you will find there or perhaps some other ones you may use refer to LAMMPS documentation. http://lammps.sandia.gov/doc/Section_commands.html

Running simulations

To run a simulation you will also need several parameter files. Most of those files you can find in awsemmd/parameters/ directory. Bellow is a short description of the files located there

  • fix_backbone_coeff.data - the main parameter file for AWSEM potential. Each term has its own section within fix_backbone_coeff.data entitled with a term name enclosed in [] brackets. To switch any of the terms off put a "-" after [] brackets (or alternat its title in any other way). [Epsilon] and [ABC] sections must be always on. [Epsilon] sets the general energy scale of AWSEM potential. [ABC] coefficients are used to calculate the positions of implicit backbone atoms from CA and O coordinates.
  • gamma.dat - contains gamma parameters for direct or protein/water mediated potential ([Water])
  • burial_gamma.dat - contains gammas for burial potential
  • para_one, para_HB, anti_one, anti_HB, anti_NHB - coefficients for beta hydrogen bonding potential ([Dssp_Hdrgn])
  • amh-go.gamma - example of gamma file used by AMH-Go term
  • uniform.gamma, alpha.gamma - sample gamma files for Fragment Memory term

You may also need ssweight file if [SSWeight] section in fix_backbone_coeff.data in on. This file is typically used to apply a secondary structure bias based on a bioinformatic prediction. This file should contain two columns of float numbers; one line for each residue indicating a probability of the residue forming alpha-helix or beta-sheet. GenSswight.py script from awsemmd/tools/create_project_tools/ can be used to convert a prediction from JPRED online server (http://www.compbio.dundee.ac.uk/www-jpred/) to the desired format. The input file for GenSswight.py should contain only one line with a prediction string.

After obtaining all necessary parameter files you can use the lmp_serial < your_project.in command to run a LAMMPS-AWSEM-MD simulation.