Re: [AMBER] unwrap problem in the liquid water dynamics trajectory analysis

From: David A Case <>
Date: Mon, 17 Sep 2018 01:59:18 +0100

On Sun, Sep 16, 2018, jinfeng liu wrote:
> Recently I performed liquid water dynamics under the periodic
> boundary condition in the NVT ensemble using the "iwrap=1". The unit cell
> of water box contains 128 SPC/FW water molecules. I need to remove the
> effect of "iwrap=1" in the trajectory before the diffusion coefficient is
> calculated.

To add to the good comments earlier:

1. I'm a little surprised you can even do a calculation with just 128
water molecules. You would certainly need more to avoid bad effects
from artificial periodicity. What is the size of your periodic box?

2. For diffusion (and for other dynamic parameters) you need to run NVE
simulations (following equilibration in NPT). Thermostats inevitably
interrupt Newton's equations, and destroy the very time-dependent
behavior you are trying to measure.

3. Why use iwrap=1 at all?

4. As noted earlier, be sure you are using SPC/FW correctly. You need
to add "set default FlexibleWater on" in your leap input.

...good luck...dac

AMBER mailing list
Received on Sun Sep 16 2018 - 18:00:02 PDT
Custom Search