CHARMM and MM(PB/GB)SA
PB model is recommended when working with CHARMMff files. Nevertheless, the combination of PB/GB models with radii optimized for amber atom types (i.e. bondi, mbondi, mbondi2, mbondi3) and CHARMM force field hasn't been tested extensively. Please, check this thread for more information and proceed with caution.
in gmx_MMPBSA v1.5.0!!!
In gmx_MMPBSA v1.5.0 we have added a new PB radii set named charmm_radii. This radii set should be used only with systems prepared with CHARMM force fields. The atomic radii set for Poisson-Boltzmann calculations has been derived from average solvent electrostatic charge distribution with explicit solvent. The accuracy has been tested with free energy perturbation with explicit solvent ref.. Most of the values were taken from a *radii.str file used in PBEQ Solver in charmm-gui.
- Radii for protein atoms in 20 standard amino acids from Nina, Belogv, and Roux
- Radii for nucleic acid atoms (RNA and DNA) from Banavali and Roux
- Halogens and other atoms from Fortuna and Costa
Protein-ligand embedded in membrane binding free energy calculations (Single Trajectory method) with CHARMMff files¶
Info
This example can be found in the examples/Protein_membrane_CHARMMff directory in the repository folder. If you didn't use gmx_MMPBSA_test before, use downgit to download the specific folder from gmx_MMPBSA GitHub repository.
Requirements¶
In this case, gmx_MMPBSA
requires:
Input File required | Required | Type | Description |
---|---|---|---|
Input parameters file | in | input file containing all the specifications regarding the type of calculation that is going to be performed | |
The MD Structure+mass(db) file | tpr pdb | Structure file containing the system coordinates | |
Receptor and ligand group | integers | Receptor and ligand group numbers in the index file | |
A trajectory file | xtc pdb trr | final GROMACS MD trajectory, fitted and with no pbc. | |
A topology file | top | take into account that *.itp files belonging to the topology file should be also present in the folder | |
A Reference Structure file | pdb | Complex reference structure file (without hydrogens) with the desired assignment of chain ID and residue numbers |
-> Must be defined -- -> Optional, but recommended -- -> Optional
See a detailed list of all the flags in gmx_MMPBSA command line here
Command-line¶
That being said, once you are in the folder containing all files, the command-line will be as follows:
gmx_MMPBSA -O -i mmpbsa.in -cs com.pdb -ci index.ndx -cg 6 5 -ct md.xtc -cp topol.top -o FINAL_RESULTS_MMPBSA.dat -eo FINAL_RESULTS_MMPBSA.csv
mpirun -np 2 gmx_MMPBSA -O -i mmpbsa.in -cs com.pdb -ci index.ndx -cg 6 5 -ct md.xtc -cp topol.top -o FINAL_RESULTS_MMPBSA.dat -eo FINAL_RESULTS_MMPBSA.csv
gmx_MMPBSA_test -t 11
where the mmpbsa.in
input file, is a text file containing the following lines:
Remember
radiopt = 0
is recommended which means using radii from the prmtop
file
See a detailed list of all the options in gmx_MMPBSA
input file here as well as several examples
Considerations¶
This is a protein-protein complex system that contains several glycosilation sites and ligands bound. All of that, is embedded in POPC:CHOL a (4:1) membrane. As you will see, gmx_MMPBSA is able to handle successfully such a complex system. Of note, just a relevant part of the entire system has been considered for binding free calculations, since the inclusion of the rest will increase the computation time without improving the results. You can check the file _GMXMMPBSA_COM_FIXED.pdb during the calculation to see how the complex looks like. In this case, a single trajectory (ST) approximation is followed, which means the receptor and ligand structures and trajectories will be obtained from that of the complex. To do so, an MD Structure+mass(db) file (com.pdb
), an index file (index.ndx
), a trajectory file (md.xtc
), and both the receptor and ligand group numbers in the index file (6 5
) are needed. The mmpbsa.in
input file will contain all the parameters needed for the MM/PB(GB)SA calculation. A topology file is also needed (mandatory) in this case to generate the topology files in amber format with all the terms for CHARMM force field.
Comments on parameters for implicit membranes
The inclusion of an implicit membrane region in implicit solvation calculations is enabled by setting memopt
to 1 (default value is 0, for off). The membrane will extend the solute dielectric region to include a slab-like planar region of uniform dielectric constant running parallel to the xy plane. The dielectric constant can be controlled using emem
. We set the membrane interior dielectric constant to a value of 7.0 in this example. The value of emem
should always be set to a value greater than or equal to indi
(solute dielectric constant, 1 in this example) and less than exdi
(solvent dielectric constant, 80.0 default).
The thickness is controlled by the mthick
option (40 Å in this case). The center of the membrane region is controlled with mctrdz
and in this case the membrane region will be centered at 37 Å from the center of the protein. If calculations are performed on a protein with a solvent-filled channel region, this region would be identified automatically by setting poretype=1
.
When using the implicit membrane model, the default sasopt=0
, i.e. the classical solvent excluded surface, is recommended due to its better numerical behavior. When running with the default options, the program will compute solvent excluded surfaces both with the water probe (prbrad=1.40
by default) and the membrane probe (mprob=2.70
by default). This setting was found to be consistent with the explicit solvent MD simulations.
It is also suggested that periodic boundary conditions be used to avoid unphysical edge effects. This is supported in all linear solvers. In the following, Geometric multigrid is chosen (solvopt=2
) with ipb=1
and bcopt=10
. The linit=1000
should work fine, but take into account that working with linear and periodic boundary conditions could require more iterations.
In addition, eneopt
needs to be set to 1 because the charge-view method (eneopt = 2
) is not supported for this application. When eneopt=1
, the total electrostatic energy and forces will be computed with the particle-particle particle-mesh (P3M) procedure outlined in Lu and Luo.[8] In doing so, energy term EPB
in the output file is set to zero, while EEL
term includes both the reaction field energy (EPB
) and the Coulombic energy (EEL
). The van der Waals energy is computed along with the particle-particle portion of the Coulombic energy. This option requires a nonzero CUTNB (in this case, cutnb=8.0
). It's noteworthy mentioning that ΔGGAS
and ΔGSOLV
as reported are no longer properly decomposed. Since EPB
and EEL
are combined into the "gas phase" term, the gas and solvation terms can't be separated. Nevertheless, the total ΔTOTAL should be perfectly fine, since everything is sum up together in the end.
Danger
Note that a smaller fillratio=1.25
is used compared to the defult one (4.0). The use of a periodic boundary also allowed a somewhat small fill ratio (i.e., the ratio of the finite-difference box dimension over the solute dimension) of 1.25 to be used in these calculations (ref). Be cautious when changing this parameter as its increase may lead to a considerable RAM usage (specially when running the program in parralel).
A plain text output file with all the statistics (default: FINAL_RESULTS_MMPBSA.dat
) and a CSV-format output file containing all energy terms for every frame in every calculation will be saved. The file name in '-eo' flag will be forced to end in [.csv] (FINAL_RESULTS_MMPBSA.csv
in this case). This file is only written when specified on the command-line.
Note
Once the calculation is done, the results can be analyzed in gmx_MMPBSA_ana
(if -nogui
flag was not used in the command-line). Please, check the gmx_MMPBSA_ana section for more information
Created: October 17, 2020 22:35:03