Hello all,
As far as I know, minimization is done before simulations but are there cases when the simulated structure still needs further refinement?
MD simulation is a only a way to solve the equations of motion for the particles in a system. These equations can be derived by minimizing the action of the system. Thus, a minimization procedure is already implicit in using the equations of motion. There is no mathematical or physical justification for minimization of any sort unless that is required by some additional facts such as the minimization of some relevant free energy. Your question does not provide any clue as to what you want to use MD simulation for so it is hard to answer it. Provide more details so that the question can be answered.
If you use simulation as the first technique, you will need to optimize (max or min) afterwards since simulation is only a performance technique. Analyze results of simulation with benefit/cost analysis or multiobjective decision making. If you use optimization technique first (for example deterministic model) you can simulate afterwards in order to incorporate stochastic aspects. Yours, Ricardo Harboe, Ph.D. in Engineering (UCLA,1970).
Hi,
It highly depends on what your are trying to get out of simulation. If the property you are looking for is dynamic (like diffusion coefficient), so you need all of the fluctuations in the system and all those frames that you have generated in the MD. But if you are looking for some phase transformation or separation, there is no harm in a short energy minimization to increase the resolution of the last frame (e.g. for better picks in RDF). But generally I never heard of energy minimization after MD. In the case of Ramachandran plot, you are interested in dynamical changes in dihedral angles respective to each other, so I think there is no need for energy minimization.
Shortly speaking, Akanksha, it depends on what do You want. But in any case You must be aware, that structures may differ at different levels of calculations.
Regards,
Eugene.
MD simulation is a only a way to solve the equations of motion for the particles in a system. These equations can be derived by minimizing the action of the system. Thus, a minimization procedure is already implicit in using the equations of motion. There is no mathematical or physical justification for minimization of any sort unless that is required by some additional facts such as the minimization of some relevant free energy. Your question does not provide any clue as to what you want to use MD simulation for so it is hard to answer it. Provide more details so that the question can be answered.
It depends on what you want to do. For instance, in order to obtain the lowest energy structure, one must minimize the sampled geometries from the MD. If you dont care about the temperature or the pressure conditions of the geometry, you may use it as is.
Optimization implies that the you will get the relaxed geometry as it will occur at zero temperature, zero pressure.
So, the bottmline is, that it is for you to decide whether or not you need to optimize the system after performing MD on it.
- Badges
- Science topic
- Similar topics
- Computer Communications (Networks)
- Simulators