While performing an MD simulation using X force field potential energy functional form (CHARMM/AMBER/OPLS), is it necessary to use force field parameters corresponding to the same X force field? Or, is it permitted to use parameters corresponding to different force field Y, or a combination of parameters from different force fields Y and Z?
Reason for asking the question is, I am working with an ionic liquid system, and AMBER parameters are available for cation, while OPLS-AA parameters are available for anion. Can I use this combination and perform MD simulation by implementing CHARMM force field functional form in NAMD?
It is not absolutely forbidden to mix parameters from different force fields, but you must be very careful. Different packages often use different definitions for force field descriptions, e.g. for Lennard Jones parameters in combination with smoothing functions, or in dihedral angle definitions. They may define molecular topologies in different ways. In such cases parameters are not interchangeable!
In your case the way electrostatic interactions are treated will be crucial. The various possibilities within one package (Ewald sum, reaction field,..) will differ more between each other than between packages.
The golden rule is: validate your method+model+parameters by checking on a sensitive experimentally verifiable property of a simple, but comparable system.
No. Mixing different parameter sets is not a valid approach. You need a self-consistent set of parameters to properly carry out a simulation.
In the publication cited here, AMBER parameters for bmim+ cation and OPLS-AA parameters for tf2n- anion were used to perform MD simulation using AMBER force field. How is it valid?
Article The [BMI][Tf 2 N] Ionic Liquid/Water Binary System: A Molecu...
AMBER and OPLS-AA share a common lineage, so there may be some compatibility among the parameters, but this approach is somewhat strange to me. I only took a quick look at the paper and its referenced parameters, but I would be very wary of any sort of approach like this based on my own experience. Perhaps it managed to "work" insofar as the goals of a particular study, but in general, papers that mix and match disparate parameter sets should be rejected. Self-consistency is crucial in force field development.
The answer depends on the task and the required accuracy.
Since you're talking about ionic liquids, it is likely that the dominant contribution will be electrostatics. In Amber and Opls it defined in very different ways. Therefore, at least you need to determine the electrostatics in a consistent manner. Everything else depends on the task.
On the other hand, against the background of integer charges, one can hardly hope on the precise definition of partial charges. So it may be that this approach is justified. It should be checked for a specific task.
It is not absolutely forbidden to mix parameters from different force fields, but you must be very careful. Different packages often use different definitions for force field descriptions, e.g. for Lennard Jones parameters in combination with smoothing functions, or in dihedral angle definitions. They may define molecular topologies in different ways. In such cases parameters are not interchangeable!
In your case the way electrostatic interactions are treated will be crucial. The various possibilities within one package (Ewald sum, reaction field,..) will differ more between each other than between packages.
The golden rule is: validate your method+model+parameters by checking on a sensitive experimentally verifiable property of a simple, but comparable system.
- Badges
- Science topic
- Similar topics
- Chemistry
- Theoretical Chemistry
- Computational Chemistry
- Molecular Dynamics