How gmx_MMPBSA works¶
gmx_MMPBSA is a new tool based on AMBER's MMPBSA.py aiming to perform end-state free energy calculations with GROMACS files.
But, what does that mean?
Basically, gmx_MMPBSA
provides all the MMPBSA.py functionalities and more to GROMACS users.
MMPBSA.py is an excellent and well-known tool to perform end-point binding free energy calculations in AMBER (January/2022 more than 1500 citations). On the other hand, there are tools like g_mmpbsa
that are well known within the GROMACS community (January/2022 more than 1700 citations). Interestingly, MMPBSA.py is more robust and was published first, however, g_mmpbsa
has higher number of citations. This is probably due to the fact that the GROMACS (Open source and free) community is large, while AMBER has the restriction of a paid license, or a small community with free academic license.
The use of MMPBSA.py for GROMACS users requires enormous effort to successfully complete the process. In that regard, we have decided to make our experience in this process available to the community. We have not limited ourselves only to the extent of utility itself, but added additional value. We have designed a tool gmx_MMPBSA
that allows to perform a number of calculations in an effortless way with a graphical tool gmx_MMPBSA_ana
incorporated, which we believe has great potential for the analysis of the results obtained. We also have incorporated some functionalities that aren't present in MMPBSA.py and made others more accessible to the beginner user in Amber.
gmx_MMPBSA general workflow¶
gmx_MMPBSA functioning can be divided into 3 parts as shown in figure 1. In the first part, Preparation
, the topologies and trajectories are generated, among other elements depending on the calculations, such as the mutants for the alanine/glycine scanning or the list of interacting residues during decomposition analysis. In the second part, Calculation
, the binding free energies and/or entropies are estimated using the selected models. Finally, in the last step Analysis
, the results can be analyzed by using the graphical user interface gmx_MMPBSA_ana
.
Required input files¶
Currently, gmx_MMPBSA
supports two families of force fields: Amber and CHARMM. Although these force fileds are quite similar, the generation of topologies differs. When Amber force filed was used to prepare the system(s), the topologies compatible with AMBER package can be generated from either GROMACS topologies or structures. Meanwhile, when CHARMM force field is used, AMBER's topologies are generated from GROMACS topologies exclusively. (Table 1).
Table 1: Required input files for every force field
Force field | Structure | Index | Trajectory | Topology | Reference Structure | Small Molecule Mol2 |
---|---|---|---|---|---|---|
AMBER | tpr, pdb | ndx | xtc, trr, pdb | Optional | Optional | Only if not top defined |
CHARMM | tpr, pdb | ndx | xtc, trr, pdb | Always | Optional | No |
Topology preparation¶
In this section, we will go in detail about each file and what they are used for.
GROMACS files
MD Structure+mass(db) (*.tpr, *.pdb)
- This file is used to generate the structure in pdb format of the complex with
editconf
ortrjconv
. We recommend using the*.tpr
(production*.tpr
) format. Index (*.ndx)
- This the file that contains the index of each element contained in the *.tpr file, organized as groups. This file is required for the definition of the groups corresponding to the receptor and the ligand.
Trajectory (*.xtc, *.trr, *.pdb)
- Trajectory files.
Topology (top)
- This file contains all the parameters corresponding to the force field selected during the system setup. When using a GROMACS topology,
parmed
is used for converting the topologies. This method can be useful when studying a complex systems with many elements or systems containing elements that have been parameterized independently and don't appear in standard force fields. If you use Amber force field, you can consider whether using the topology or not. However, when using CHARMM (any version) force field, the topology is always required. Reference Structure
- It corresponds to a file in *.pdb format that must contain a complete structure. That is, the user must make sure that this structure contains the same number of atoms and residues as the complex that he initially defined, the correct residue numbering, as well as its chain ID. This structure is optional, but we recommend using it since it should guarantee a smooth processing of the files in
gmx_MMPBSA
. Essentially, the objective is to be able to correctly assign the mentioned parameters since internallygmx_MMPBSA
handles sensitive information, for example: when it extracts the receptor and the ligand from the complex structure, the mutation in the alanine scan, etc. Small molecule parameters (*.mol2)
- This file contains a ligand parameterization with
antechamber
that is not found in the selected force field (Amber family). It is only necessary to define it when studying a system that contains this type of ligand, and a topology has not been defined. This is used to build Amber topology from structure usingtleap
.
Here are the steps that gmx_MMPBSA follows to generate the topologies: Specific steps for the topology are highlighted like this
- Generates a new index and registers the groups defined by the user
- Generates the pdb of the complex, receptor, and ligand
- Topologies are cleaned (remove water and ions)
- The structures or parameters for the receptor and the ligand are generated if it is an ST approximation.
- If Alanine scanning: the mutant of the complex, and the mutant receptor or ligand are generated
- If decomposition: interaction residues are extracted
- The complex is mapped. Registers the continuity of the receptor (example: Metalloprotein structure files are generally structured as follows: Receptor protein + Ligand + Receptor Ions)
- The PBRadii is assigned
- Topologies are converted with
parmed
or generated withtleap
The following figure shows the process to generate AMBER topologies depending on the force field.
Created: February 8, 2021 07:10:13