I am currently running a umbrella sampling process where the calculation of some property is highly dependent on the positions of ions separated by a protein embedded in a membrane. However I have run into a problem where my bulk solution ions are at dramatically different positions between the windows due to wrapping at the boundary. Like window 1 will have 23 chlorides on the 'top' side of the membrane, window 2 will have 17, window 3 will have 21 and so on.
I am running under 2 different conditions, condition 1 has an electric field and condition 2 does not. I have been trying to converge the no electric field bulk solution in NAMD using NPT conditions. I am currently at 0.4ns with no solution in sight. Is it possible to converge the bulk solution without a voltage applied? Are my parameters off?
I am using NAMD with a Drude polarizable force field
Conditions:
Timestep of 1.0 femtosections for 400,000 steps, = Langevin thermostat with a Langevin piston target of 1.01325 and I have specified periodic boundary conditions to be 92x92x100. Drude hardball is on with a bond length of 0.25. I am unsure if specifying a boundary and having a pressure target could be causing issues.
I know that, when I apply an electric field I will have a distribution of ions that will cancel out the effect of the electric field. meaning I should have a consistent number of CLAs between the window but I cannot seem to get the no electric field one to converge to a consistent number.
- Badges
- Science topic
- Similar topics
- Biological Science
- Bioecology
- Systems Ecology
- Landscape Ecology
- Connectivity