From: Olaf Lenz (olenz_at_icp.uni-stuttgart.de)
Date: Mon May 23 2011 - 04:57:38 CDT

Hi!

It is very hard to find your problem if I can't have a look at the
actual files that cause the problem. If you want me to help, please make
the trajectory file available somewhere, or send it to me via private
email. Please don't send it to the list, if it isn't very small.

On 05/20/2011 03:05 PM, Nicola Giacche' wrote:
> I have a desmond trajectory of 1.2ns, with several "jumps"... with
> the command "pbc unwrap" I resolved these jumps, but not for whole
> trajectory, in the last part some molecules have, again, overlong
> bonds not resolved by the unwrap command.

"pbc unwrap" will try to make the trajectory as smooth as possible by
comparing the position of an atom in a frame with the same atom's
position in the previous frame. If the atom makes a jump of more than
half a box length, it will try to find the minimum image of the new
position.
Maybe if you know how the function works, you are able to debug the
problem yourself.

> Furthermore, after the unwrap, in order to resolve the "diffusion" of
> the solvent molecules, I will apply the command "pbc wrap -all
> -center com -centersel "resname UNK" -compound res", but if I don't
> fix completely the system with the unwrap command also the
> subsequent wrap command does not resolve the "diffusion" of the
> solvent from the box.

Again, this is hard to debug if I can't see the trajectory.

Olaf

-- 
Dr. rer. nat. Olaf Lenz
Institut für Computerphysik, Pfaffenwaldring 27, D-70569 Stuttgart
Phone: +49-711-685-63607