search

openmm / openmmforcefields

CHARMM and AMBER forcefields for OpenMM (with small molecule support)
http://openmm.org
Other
248 stars 80 forks source link

<< OpenMM forcefield for organic/inorganic systems >> #241

Open icamps opened 1 year ago

icamps commented 1 year ago

Hello,

I am interested in molecular dynamic simulations for systems other than proteins/membranes, etc.

Specifically, I want to run simulations for system like carbon nanostructures or boron-nitrogen nanostructures interacting with metals and green-house effect gases.

Does OpenMM has any forcefield available for those systems?

Thanks.

peastman commented 1 year ago

None of the current force fields supports that. It's a bit outside my field. Are there standard force fields that people use for those systems?

mattwthompson commented 1 year ago

My experience with some systems like these (2D materials, gas adsorption, sp2 carbon with defects) as a grad student (2013-2018 or so) was that there are no standard force fields for this sort of screening work and people generally used whatever could produce parameters, whether or not they were sensible for the application being studied. People generally used GAFF/OPLS-AA/etc outside of its range of applicability and/or made one-off force fields for the systems they're studying.

Taking a force field off the shelf and applying it to these systems is probably not going to yield great results for interesting systems, in my opinion. The easiest way to get stuff working here (without fitting new force fields) is probably searching through the literature for particular parameters that have been used and writing them up into new force field XMLs.

icamps commented 1 year ago

Thank you very much @peastman and @mattwthompson for your answers.

There are force fields like Universal force field (UFF) and Dreiding force field that have "generic" parameters. Of course that the results will not be so accurate as specific force fields...

Also, the reactive force-field (ReaxFF) has parameters for many elements from the periodic table.

jchodera commented 1 year ago

These force fields would certainly be neat to add to OpenMM, if even for simple experimentation or educational purposes. OpenMM is certainly capable of simulating parameterized systems. Someone would need to craft a Python script to assign the parameters and create the System object (or an OpenFF Interchange object) along with associated tests, but this isn't something we have had many requests for.

Perhaps this could be labeled a feature request?