<< Membrane protein PSF/DCD >>

[Ropón-Palacios G.]

Hello,

I want to perform a MMPBSA calculation for a system that contains membrane+protein+ligand, but my systems have PSF and DCD format, could you give me an example of how to do this, and the ` *.in`  file for both this calculation and the one for descomposition analysis

?

Geo.

—
Ropón-Palacios G.

Computational biologist & biophysicist,

Laboratorio de Modelado Computacional, 

Intituto de Ciencias Exactas, 

Universidad Federal de Alfenas, MG, Brasil.

Phone: +51 935 055240.

Principal e-mail: gro...@gmail.com

 

[Mario Sergio Valdes]

Welcome Geo.

To convert your PSF and DCD files, you can check these examples (https://valdes-tresanco-ms.github.io/gmx_MMPBSA/dev/examples/#support-for-psf_dcd-files). Currently, we have limited support for these file formats, but we are working in xBFreE, which has native support for GORMACS, AMBER, NAMD, and CHARMM programs.

To perform the BFE calculation, including membranes, you can check these examples (https://valdes-tresanco-ms.github.io/gmx_MMPBSA/dev/examples/Protein_membrane/, https://valdes-tresanco-ms.github.io/gmx_MMPBSA/dev/examples/Protein_membrane_CHARMMff/)

As the energy decomposition is an extra analysis, you can simply add the decomp name list to your input file. Please, check this example (https://valdes-tresanco-ms.github.io/gmx_MMPBSA/dev/examples/Decomposition_analysis/).

Important points:

- Please, check the documentation carefully. The membrane approximation requires defining several parámeters according to your system

- gmx_MMPBSA only supports all atoms ff, excluding OPLS with impropers, unified atoms style, or GROMOS style

- Please, take a look at the parámeters of energy decomposition calculation since in this context the default value of print_res can select a lot of residues

Any errors or doubts contact us.

Happy BFE calculation

[Ropón-Palacios G.]

Hello, in the membrane protein example, can I use it with namd PSF/DCD trajectories or is some special treatment required? And how are the indexes generated? It would have to be protein+membrane (index 1) and ligand (index 13), for say.?

[Mario Sergio Valdes]

Yes, you can use it as well. The index file is generated by gromacs using a structure file that map the structure according to its components. Following the tutorial, you will able to generate it.

No, the membrane is not needed since this approach simulates the membrane's electrostatic environment, so your groups should contain protein (receptor) and ligand

[Ropón-Palacios G.]

Great, my question now would be a few:

1. Can the trajectory and the PDB only have the ligand and the protein?, to eliminate the other components and make the system lighter?

2. The option that mentions `poretype=1` only applies if there are water molecules, even if they are transient? Or is it strict for those channels that transport water or with that capacity?

3. Can the `mctrdz=37` option be the center of mass of the membrane in Z?

4. In the `mthick=40` option, is this a measure from phosphate to phosphate, and were oxygen cholesterol also taken into account? Or is there any recommendation for this, if I have more complex mixtures.

Thank you.

[Mario Sergio Valdes]

1. Can the trajectory and the PDB only have the ligand and the protein?, to eliminate the other components and make the system lighter?

Not necessarily. The variable  solvated_trajectory defines if the trajectory needs to be cleaned or not. It means that if you solvated_trajectory=1 (default), gmx_MMPBSA will generate a cleaned trajectory for you containing only the groups you defined

 

2. The option that mentions `poretype=1` only applies if there are water molecules, even if they are transient? Or is it strict for those channels that transport water or with that capacity?

This variable defines if a pore will be identified if exist. At this time, I really unknown if this method is applied per frame or if it makes exploration in the md to identify if any pore exists. I think that is a per-frame analysis, so it will identify the pore when it exists.

 

3. Can the `mctrdz=37` option be the center of mass of the membrane in Z?

Oh, sorry about this misinformation. According to the Amber manual, this variable defines the "Absolute membrane center in the z-direction (Default=0.0, use protein center as the membrane center)". But I prefer to define it because the protein geometric/mass center not necessarily is the center of the membrane. In this case, it is the geometric center.
protein.

 

4. In the `mthick=40` option, is this a measure from phosphate to phosphate, and were oxygen cholesterol also taken into account? Or is there any recommendation for this, if I have more complex mixtures.

 

I think this trivial definition does not generate a problem o large error. You can define it as you believe It has more sense because there is no clear definition. I mean, I haven't too much experience in this type of calculation, but 3 o 4 A more or less there shouldn't be a big effect. The default value cover almost every type of membrane 

Let me know any other questions.