rahagr / freeenergy

This repository is a bash script to prepare the simulation files for FreeEnergy calculation in GOMC.

FreeEnergy

This repository is a bash script to prepare the simulation files for absolute FreeEnergy calculation using TI and FEP methods in GOMC. GOMC currently couples LJ and Coulombic interaction with lambda, using soft-core or hard-core scheme, as implemented in GROMACS. For more details regarding the coupling scheme, please refer to GROMACS Documentation.

Software dependency:

1. packmol for packing molecule in box.
   • Update the packmol executable file in build/pack/. directory.
2. vmd for generating PSF file.
   • vmd executable must be added to the path.
3. GOMC Monte Carlo code to run simulation in NVT and NPT ensemble.
   • Update the GOMC executable files (GOMC_CPU_NVT, and GOMC_CPU_NPT) in build/input/. directory.
4. alchemlyb for analysis the results using TI and FEP methods.

What does it do?

The build.sh will pack the box with defined number of solvent and a solute, generate PDB and PSF file for solvent and solute. It creates directory for each solvent. Inside each solvent directory, it creates EQ direcotry for NVT and NPT equilibration simulation. Then it creates TI_ direcotry for each replica and inside that create separate state_0, state_1, ... directory.

There are two python scripts available to calculate the free energy using BAR, MBAR, and TI estimators in alchemlyb python library.

1. freeEn_correlated_data.py: Calculate the free energy difference using correlated samples.
2. freeEn_uncorrelated_data.py: Calculate the free energy difference using uncorrelated samples.

Solvent
|___________ build
|            |______ input files (config files, GOMC executables, job scripts)
|            |______ pdb files (single PDB files for solvent and solute)
|            |______ model (parameter and topology files for solvent and solute)
|            |______ pack (initial PDB and PSF files for equilibration)
|
|___________ EQ
|            |_______ NVT (running NVT simulation prior NPT, using PDB and PSF files in build/pack)
|            |_______ NPT (running NPT simulation from equilibrated NVT simulation, using PDB and PSF files in EQ/NVT)
|
|____________ TI_1
             |_______ state_0 (equilibrte the system with NVT simulation with lambda state 0, using PDB and PSF files in EQ/NPT) 
             |                (production run in NVT simulation with lambda state 0, using equilibrated system)
             |_______ state_1 (running NVT simulation with lambda state 1, using PDB and PSF files in EQ/NPT)
             |                (production run in NVT simulation with lambda state 1, using equilibrated system)
             |_______  ...

How to modify the configuration file

To modify the configuration file, modify the NVT.conf(for NVT equilibratio), NPT.conf (for NPT equilibration), eq.conf (NVT equilibration with lambda i), and prod.conf (production run in NVT) in build/input/ directory.

How to add another solvent or solute

To add another molecule, you need to:

1. Generate the PDB file for single molecule in build/pdb/ directory. (using gaussview or VMD molfacture tools)
2. Update the Topology.top file in build/model/
3. Update the parameter.par file in build/model/

How to modify the system parameter?

Free energy systems:

We can define a range of solute name in build.sh to calculate the free energy.

#system info
NREP=5                                                      # number of TI replica
SOLVENT="octanol.pdb"                                       # pdb name of solvent
TOT_SOLUTE=(pfmethanol.pdb pfethanol.pdb pfpropanol.pdb)    # pdb names of solute 
TOT_SOLUTE_RESNAME=(F3O F5O F7O)                            # residue name of solute 

Config file:

There are some options in config files that can be controls by build.sh:

#simulation info
T=298                                            # temperature (K)
P=1.01325                                        # pressure (bar)
RCUT=14                                          # cutoff distance (A)
RCUTLOW=0.00                                     # hard cutoff (A) to avoid overlap
FE_MC=50000000                                   # steps for free energy simulation
EQ_MC=2000000                                    # steps for equilibration simulation (both NVT.conf and eq.conf)
NPT_MC=10000000                                  # steps for NPT equilibration simulation
FE_FREQ=5000                                     # free energy calc frequency
STATES=(0    1    2    3    4    5    6    7)    # state id (for user only) 
VDW=(0.00  0.20 0.40 0.60 0.80 1.00 1.00 1.00)   # lambda vector for VDW
COUL=(0.00 0.00 0.00 0.20 0.40 0.60 0.80 1.00)   # lambda vector for Coulomb

Packing parameters:

We can set the number of solvent and box size to be packed in build.sh:

#packing info 
N_SOLVENT=200                                    # number of solvent
BOX_SIZE=37.6                                    # length of cubic box (A)

Parallelization:

Modify the job script job.sh in build/input/. We can set the number threads for openmp parallelization in build.sh:

#packing info 
CPU=4                                            # number of CPU threads

How to run it?

Change your directory to this repository and execute the following commands in your terminal:

cd ~/Desktop/FreeEnergy
./build.sh

If you are dealing with job scheduler, execute to following command to run the equilibration simulations:

./eq_subjob.sh

Once the equilibration simulaitons were finished, execute to following command to run the production simulations:

./pr_subjob.sh