How to set path of an external gromacs in conda medium?

[Alisher Eshimbetov]

Dear all,

I have simulated a protein-ligand complex using GROMACS with the amber99sb-ildn force field. A KeyError (NT) occurred in MMPBSA, and it turns out that MMPBSA uses GROMACS version 5.x. Therefore, I would like to use my GPU/GROMACS in a Conda environment to determine the protein-ligand binding energy using MMPBSA.

How can I set the path for an external GROMACS installation in a Conda environment? Alternatively, how can I update GROMACS to version 5.x in my Conda environment?

Thanks in advance,
Alisher

[mariosergi...@gmail.com]

gmx_MMPBSA uses any gromacas > 4.x. You see "[INFO   ] Using GROMACS version > 5.x.x!" because the execution statement changed with the 5.x.x version, for example, `make_ndx` to `gmx make_ndx`.
So, you can use any version of gromacs you want (we recommend using the latest stable versions).

I can't identify the error. Could you please attach the gmx_MMPBSA.log file for review?
If you still need to use another gromacs different from the environment or PATH, you can configure the external path using the gmx_path variable in the input file.

[Alisher Eshimbetov]

I have prepared topology by acpype and during the simulation i have added missing NT atom to the list in the ffnonbonded.top file.

[mariosergi...@gmail.com]

Why did you add them?
I can't quite understand the problem. Please send me the files you used (only 5 XTC frames) to reproduce the error.

[Alisher Eshimbetov]

Dear Sir!

I reinstalled gmx_MMPBSA and got the error message again.

[mariosergi...@gmail.com]

As you can see, UNL_GMX.itp contains a section for the `[ atomtypes ]` section, which includes the NT type. If you keep this section in the UNL.itp file, then the calculation works as expected.
Please, don't edit any file in the ff folder as least you are sure what you are doing. In this case, fortunately, the parameters are the same from acpype and the ff files, but this can change.

[Alisher Eshimbetov]

Dear Sir,

  Thank you so much! 

According to your instructions, I kept the atomtype in the .itp file, and the GMX MMPBSA calculation is complete