solvation energy calculation with customized force field parameters in gromacs format

[Chungwen Liang]

Dear developers,

I would like to perform only the solvation energy calculation (estimation) for a given molecule in the gas phase. Would gmx_MMPBSA work for this purpose? 

If so, how to link gmx_MMPBSA to an existing force field folder (in-house developed) to generate topology in the amber format (to perform gmx_MMPBSA)? Thanks very much for the answer. 

Best,

Chungwen

[Mario Sergio Valdes]

Interestingly, although gmx_MMPBSA includes desolvation energy calculation, it has no way of excluding the rest of the terms (MM and SA) calculation. I think I'll put it on the to-do list. In this case, you will have to calculate everything, and then extract the information you need. You can select PB, GB, GBNSR6 and RISM and their derivates for the calculation.
As for the ff, that depends. If your custom force field is based on Amber or Charmm, you should have no problems using the topology with the -cp topol.top option on the command line. If it is based on unsupported force fields such as GROMOS, OLPS-UA, or any others that have additional definitions such as improper angles, or unified atom models, you will not be able to use them.

[Chungwen Liang]

Thanks for the feedback. I was trying the 3D-RISM method and the output does show each individual energy term (BOND/ANGLE ... and GSOLV). 

I would guess the GSOLV term is what i want. 

However, is there any way to set up the calculation only for a model compound as solute and calculate its solvation energy? 

would a system (complex) containing a model compound (receptor) and a dummy atom (ligand) work? Thanks again.

[Mario Sergio Valdes]

I don't quite understand what you mean. Do you want to compute the solvation energy of the ligand only? If so, then the best option would be to run it manually... Keep in mind that 3D-RISM takes much more time than PB and GB

[Chungwen Liang]

yes, I would like to calculate the solvation energy of the ligand only. Is there any tutorial/test case/web page demonstrates how to run the code manually that I could refer to? Thanks 

[Mario Sergio Valdes]

Well, it depends on what method you want to use. 

Amber (sander) support PB, GB, 3D-RISM. Also, you can use gbnsr6 and rism3d.snglpnt for calculations using the Amber topology file. pbsa or pbsa.cuda allows you the calculations using a topology file as well as a pqr file.

There are some servers for the PB calculation, for example, APBS (https://server.poissonboltzmann.org/) or DelPhi (http://compbio.clemson.edu/sapp/delphi_webserver/) which also use the pqr files as input

Unless you want to run a lot of ligands, servers are the best option