Let's say I want to run a simulation for benzene in GROMACS using parameters from a certain force-field. I logged on to the Automated Topology Builder, I looked up benzene, and I found an ITP file for benzene, and a structure (pdb) file for benzene.

In the ITP file, they mention:

"bonded and van der Waals parameters were taken from GROMOS54A7 parameter set."

This sentence makes me think that the parameters they have provided can only be used when I am using the GROMOS54A7 forcefield.

So my topology file has the statement #include GROMOS54A7_ATB.ff.

My question is, what if I want to run a simulation for benzene in CHARMM or OPLSS or some other forcefield in GROMACS? Do I have to manually construct the .itp file, with the right atom types, and everything? Or is there a database where I can get the data from to use in my simulation?

  • $\begingroup$ For CHARMM specifically, you can use paramchem.org to automate this search. This doesn’t entirely answer everything in the question, hence why it’s a comment. $\endgroup$ Commented Feb 15, 2021 at 17:26
  • $\begingroup$ Thank you for your response @kskinnerx16. I am on CGENFF. I have a pdb of benzene. How do I go about getting the CHARMM parameters? $\endgroup$
    – megamence
    Commented Feb 15, 2021 at 17:32
  • 1
    $\begingroup$ You’ll have to register for an account first, and then you can use the “Upload molecule” tab. You’ll need to submit it in .mol2 format, which can be obtained using Open Babel. This can be done on the command line or at: webqc.org/molecularformatsconverter.php $\endgroup$ Commented Feb 15, 2021 at 17:35
  • $\begingroup$ @kskinnerx16, i tried your suggestion. I am getting an error when I give CGENFF my mol2 file... $\endgroup$
    – megamence
    Commented Feb 15, 2021 at 20:26
  • $\begingroup$ I’m not sure what the error is, but it’s the description of the error is usually pretty good. $\endgroup$ Commented Feb 15, 2021 at 23:28

2 Answers 2


Running simulations with small molecules is one thing that I found quite difficult especially because most online guides/tutorials are focused on proteins. So, here's a rough idea of what to do:

Getting forcefield parameters: Your question is about this, so I will give a detailed explanation. There are many forcefields to choose from, e.g. MMFF94, UFF, CgenFF, GAFF, OPLS-AA etc. Which one you use depends on what you are trying to study. Some general forcefields have been parameterized to be compatible with protein force fields, for example CgenFF with CHARMM36, GAFF with Amber as you can probably guess. I am by no means an expert on this topic, so I will advise you to look through the literature for which force field (or combination) will suit your purpose the best.

Getting parameters in a gromacs compatible format can also be tricky. There is a guide on how to do that with some popular forcefields here. For the sake of being a complete answer, I will summarize the information briefly:

  • GAFF - You can use Antechamber with the python module acpype (which has been mentioned in the other answer) to generate gromacs compatible topology files.
  • CHARMM - For this you can use the CgenFF official website (https://cgenff.umaryland.edu/) or as it was known earlier, paramchem.org. This site produces parameters in the form of .str files which have to converted to gromacs compatible format using the cgenff_charmm2gmx.py python script which is available here.
  • GROMOS - ATB can generate gromacs compatible files.
  • OPLS-AA - The LigParGen server from Jorgensen group produces gromacs compatible files.

There is also another server called swissparam which can generate parameters for small molecules, however it takes most of the bonding and charge parameters from Merck Molecular Force Field, and Van der Waals parameters from CHARMM22. The parameters are supposed to be compatible with CHARMM protein forcefield, but I have never used it with a protein so I can't be sure.

Running the simulations: In the question you have written that you already have a topology file. For other readers of this answer, a standard gromacs topology file would look like this:

; Include forcefield parameters
#include "charmm36-jul2020.ff/forcefield.itp"
; Include small molecule parameters
#include "ligand.itp"

; Include TIP3P water topology
#include "charmm36-jul2020.ff/tip3p.itp"

;Include topology for ions
#include "charmm36-jul2020.ff/ions.itp"

[ system ]
small molecules

[ molecules ]
; Compound        #mols
LIG                 1

Here I have named the small molecule (in your case benzene) as LIG, and its parameters are stored in the ligand.itp file (which you can get from swissparam, or after converting the str file from CgenFF website).

Also note that you don't always need to include the original forcefield parameter files as you have done in the question. If the .itp files say that the "parameters were taken" that means they are already present in the .itp files. However, different forcefields use different combining rules, which are defined in the forcefield.itp for that forcefield, so it has to be included. In my case, I included the ion and tip3p water topolgy because I needed them later. The forcefield.itp is also required for the tip3p water parameters.

After you have the topology and parameters, you can then run gromacs as usual.

  • 1
    $\begingroup$ @ShoubhuRMaiti, I used LigParGen to get an ITP and GRO file for benzene. My question is, can I define a topol.top file and use the inbuilt oplsaa.ff forcefield in GROMACS? the ligpargen used atomtype oplsaa_804 which is an atomtype i did not find in the aminoacid.rtp file in my oplsaa.ff directory $\endgroup$
    – megamence
    Commented Feb 25, 2021 at 18:19
  • $\begingroup$ @megamence I think the atomtype names are linked to the atom serial number and residue name in the .itp file itself, in the [atoms] section of the .itp file. All the parameters and atom definition for the benzene are self-contained in the .gro and .itp files that you got. I think the only thing that the oplsaa.ff/forcefield.itp provides is the combination rule for non-bonded terms. So, yes, I believe you can use the inbuilt oplsaa.ff forcefield without any problems. Still, I would advise you to run a short check run to see if everything is behaving correctly. $\endgroup$
    – S R Maiti
    Commented Feb 25, 2021 at 20:27
  • 1
    $\begingroup$ Hi @ShoubhikRMaiti. I am having problems with running my benzene simulation on GROMACS. It would be a huge help if you could look into it. Thank you so much again, mattermodeling.stackexchange.com/questions/4436/… $\endgroup$
    – megamence
    Commented Mar 3, 2021 at 4:45

You might want to have a look at acpype which is compatible with GAFF and AMBER-based force fields and probably others.

  • 2
    $\begingroup$ This is probably better suited as a comment imho $\endgroup$
    – S R Maiti
    Commented Feb 19, 2021 at 23:01

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .