mesosclae/DPD simulation

Hello everyone,

I am a grad student and I have been running DPD simulations in my work using Materials Studio software. But I think it has many constraints and hence I want to switch to another software. Does anyone have a manual where I can understand how to setup DPD simulations from scratch in GROMACS? I do not have any experience using it but I think I should be able to learn it fast.



DPD (dissipative particle dynamics?) is currently not supported in GROMACS.


I was not aware of that. Thanks for letting me know. I had come across some articles which used GROMACS for coarse-grained MD so I assumed that we would be able to setup DPD simulations as well. Do you have any suggestions of software for running DPD calculations?


You are welcome.

I think the confusion is due to a linguistic ambiguity: ‘coarse graininig’ molecular dynamics doesn’t really mean performing mesoscale simulations (at least in this community). At the end of the day, coarse-grained MD is just MD, but with less degrees of freedom.

Do you have any suggestions of software for running DPD calculations?

I am not really sure: the only times I run DPD simulations I used in-house codes.

I think LAMMPS has DPD (compute dpd command — LAMMPS documentation). There’s also SPARTA (, which is a direct simulation Monte Carlo code, so conceptually similar to DPD, and it’s mainly used for gas dynamics. Finally, there’s this paper:, reporting an implementation of DPD on a older version of GROMACS, you may try and ask the authors.

1 Like

Apologies my last reply (deleted) had missed a couple of the points in the above messages.

There is also DPDmacs although I’m not sure it has been used in years:

Judging from the mentioned processors, this code is nearly 2 decades old.

Yes, it isn’t even available to download anymore without requesting it.

I did try it over ten years ago and it did work back then but there must be something better out there that’s non-GROMACS (LAMMPS probably, as suggested).

LAMPPS can likely do DPD, but is often at least on order of magnitude slower than GROMACS.

Another question is if DPD is needed at all. In many cases DPD makes you run slower, but in terms of more time taken per MD step and in terms of slower dynamics, without scientific gain. DPD is only of actual use when correct hydrodynamics is required for systems with implicit solvent. This is not often the case and that is also the main reason why DPD has not been implemented in GROMACS (apart from the technical complication that the friction involves pair terms).

Maybe you can try a software called EspressoMD, it is also for DPD.