1. I have used many ways in order to prove the stability of my conformation system. One of the way is to use equilibration using canonical ensembles (NVT) followed by isothermal isobaric ensembles (NPT). In some point, I realize that there was no sole purpose to do the NVT ensembles since the system will change during the the NPT (pressure/temperature) equilibration. What do you think ?
2. How much enough simulation time needed for a simulation? What I mean is, how long the simulation will be used to analyse after it stabilizes? Depends to the system we used to study?
3. If the system didnt achieve stabilization (large RMSD value fluctuations) how should I do?
* My system is on insulin which aggregates and degrades too much. I have used many ways to let it stable but still show no good results.
Hope to get good opinion.Thank you.
The purpose of doing NVT before NPT is typically based on algorithmic stability. Velocity generation at the outset of a simulation is imperfect, and if coupled with a barostat, can frequently crash. So equilibration is often done under NVT for a short time to get the velocity distribution reasonable, followed by NPT to get the density right.
Perhaps your biggest problem here is that you have the order of NPT and NVT simulations back to front. A new simulation in explicit solvent should always start in the NPT ensemble. The reason for this is that the water is never perfectly packed - it needs to "soak in" to the protein. If you start in NVT your pressure will immediately drop through the floor and you will develop serious artifacts due to the development of voids - your simulation will, in effect, be boiling. In NPT, on the other hand, the box shrinks in size to maintain a constant pressure during this settling. Only when the box volume has converged to a stable value should you consider swapping to NVT.
... and yes, I'm afraid the answer to your question 2 is, how long is a piece of string? Every system is different, and the timescale necessary can vary over orders of magnitude depending on what you're looking for.
The purpose of doing NVT before NPT is typically based on algorithmic stability. Velocity generation at the outset of a simulation is imperfect, and if coupled with a barostat, can frequently crash. So equilibration is often done under NVT for a short time to get the velocity distribution reasonable, followed by NPT to get the density right.
1. I usually start with NPT and never saw that it led to some destabilization. (However, this depends on the type of barostat in use)
2. Consider water. I would say that the minimum protocol is 200 ps equilibration and 100 ps for a statistics. Normal protocol is 1 ns both.
3. What do you mean by stability in this case?
In addition - less of scalings, the faster the simulation. Ideal situation would be if MD could be sufficiently precise without any couplings to thermostats or barostats.
To answer your question, you need to know how "bad" you system is. For long time simulation under either NPT or NVT, MD searches for local basin on potential energy surface. If temperature is high enough, and if your initial system happens to lie in a basin next to one corresponding to "collapse" configuration, your system will go down if simulation time is enough. Usually, you need to minimize your system's energy statically by conjugate gradient method or others before you use NVT or NPT. By doing so, just imagine that, your system will go to a much much deeper basin than that for initial configuration so that the probability for your system to breakdown is largely reduced.
After energy minimization, required time for relaxing a system under NVT or NPT depends on the size and sound velocity of your system. Usually, you need to let a "elastic" wave propagates in your system for at least two times. And there is another situation, if the energy barrier between initial configuration and your expected configuration can be estimated, the required simulation time can also be estimated.
- Badges
- Science topic
- Similar topics
- Chemistry
- Theoretical Chemistry
- Computational Chemistry
- Molecular Dynamics