Hello,
I prepared my protein-ligand complex for MD run, the average value for NVT is 299.99 which looks pretty satisfactory, but for NPT, the average value is 10.01 which is strange to me. I think, this should be close to 1. Please suggest me whether this average value depends or not? Also my system is enough suitable for simulation or not.
Thanks in advance!
Dear Amir,
I would say that having fluctuation in pressure is generally expected in condensed-phase MD simulations.
See, the pressure is a macroscopic property of the system and as such, the only meaningful value of it is the average value ofter long simulations. So, the system size and timestep/total simulation time will affect the average value of the pressure. Moreover, It also depends on the type of barostat and thermostat you are using for the simulation run. So, choose it with the great care. Also, you can change the volume of the box (either increase or decrease it).
Better understanding of the thermodynamics of this problem, you can get at this link....http://ambermd.org/Questions/pressure.html
As per my understanding, we cant say anything about the right and wrong about the results rather its a combination of factors and the type of insight we want in. I also suggest you to ask this question specifically to your program's listserv, if available, to see what others have experienced.
Sir Elvis,
These are values for system equilibration, 299.99 K is constant temperature at which I want to simulate my system while 10.01 bar shows the current pressure my system has. but the standard pressure for routine simulation should be around 1 bar as Justin has mentioned in his tutorial for gromacs (http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin/gmx-tutorials/lysozyme/07_equil2.html)
Thank you sir
Dear Amir,
In principle, the MD simulation with NVT ensemble would not fluctuate the temperature much. Normally, we opt 300K as target temperature that should remain constant in whole NVT-MD run.
But while simulating the system with NPT ensemble, the philosophy of statistical mechanics is little bit changes, as you want to kept both the inter-related thermodynamic variable as constant (pressure, P and temperature,T). In such cases, the standard deviation in pressure would be large and you will eventually have large average value of pressure(i.e. 10.0 bar).
Now, coming back to my suggestion:
Most simulations are carried out in the Isothermal-Isobaric ensemble (NPT). However, to equilibrate a system correctly, it is necessary to run a short simulation in the NVT ensemble (for velocity randomization) until a suitable temperature is reached, as the calculation of pressure is inaccurate at low temperatures. thereafter, continue your NPT simulation.
Thanks Mohd Athar,
In my case, the temperature is okay, this is 299.99 K. but the NPT is giving problem. I tried different approaches like to reduce the simulation time for NPT, increasing simulation time etc, but I failed to get 1.0 bar pressure which mimics the real system. Can you please suggest any approach to solve this issue?
Regards!
Dear Amir,
I would say that having fluctuation in pressure is generally expected in condensed-phase MD simulations.
See, the pressure is a macroscopic property of the system and as such, the only meaningful value of it is the average value ofter long simulations. So, the system size and timestep/total simulation time will affect the average value of the pressure. Moreover, It also depends on the type of barostat and thermostat you are using for the simulation run. So, choose it with the great care. Also, you can change the volume of the box (either increase or decrease it).
Better understanding of the thermodynamics of this problem, you can get at this link....http://ambermd.org/Questions/pressure.html
As per my understanding, we cant say anything about the right and wrong about the results rather its a combination of factors and the type of insight we want in. I also suggest you to ask this question specifically to your program's listserv, if available, to see what others have experienced.
Hi Amir
First, there is no such hard and fast rule to prove that your system is in equilibrium. We can only check if it is not out of equilibrium.
But you can control the pressure of the system by controlling the volume of the simulation box (PV = NRT)
Hope this might help.
"Future stock prices cannot be predicted from historical stock prices."
- Badges
- Science topic