On Sat, Nov 15, 2003, franck wrote:
>
> I am interested in the calculation of the diffusion constant of the
> molecules of my system and the solvent used is water (TIP3P model at 298 K).
> For this purpose, I used the diffusion option in ptraj and followed the
> manual instruction. I got high D values for water (around 4,5) as described
> in the litterature. This value was obtained without using the image option
> of ptraj and center the whole system on my solute. But after performing
> image(center on my solute), I got a D of 2,8 cm2/s closer to the
> experimental one (2,3 cm2/s at 298 K). Nevertheless, the value is high.
>
> In both case, I first plot the MSD vs time (ps) in a log x log plot to
> identify the part of the curve to consider (diffusive regime). Then I fitted
> this curve and got the D value.
>
> I know that the MD I ran were short (2 ns), but I wonder if this high D
> value could eventually be the result of the diffusion of the water close to
> the surface of the BOX. More than 2200 water molecules are considered which
> means that 40 % of the water are in contact with the BOX walls. If one
> molecule escapes the box from one side and enters again in the box but by
> the opposite one, the calculated distance could be biased by the dimension
> of the box ? (a displacement of 2 Angstrom could be interpreted as 2 + the
> length of the box?).
>
The first value (without imaging) should be correct. This follows the path of
individual water molecules wherever (i.e. into whatever unit cell) they go.
Imaging should not be used in such a calculation (nor should iwrap=1 in
sander), for the reasons you outline above. One of the principal reasons to
allow imaging not to take place is to be able to calculate things like
diffusion constants.
...hope this helps....dac
--
==================================================================
David A. Case | e-mail: case.scripps.edu
Dept. of Molecular Biology, TPC15 | fax: +1-858-784-8896
The Scripps Research Institute | phone: +1-858-784-9768
10550 N. Torrey Pines Rd. | home page:
La Jolla CA 92037 USA | http://www.scripps.edu/case
==================================================================
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Sat Nov 15 2003 - 20:53:01 PST
