VMD-L Mailing List

From: Nicola Giacche' (nicolag_at_chimfarm.unipg.it)
Date: Tue May 24 2011 - 09:42:19 CDT

Hi,

the big "jumps" were due to "glue" feature of Desmond. Indeed in my
specific system, where there were not "solvent" but only "solute"
molecules (including the custom solvent), caused such a problem.

Turning off the "glue" feature, as recommended by James Watney of the
Desmond-User Group, the simulation runs smooth without jumps.

Nicola

Il 23/05/2011 17:20, Olaf Lenz ha scritto:
> Hi!
>
> The reason why "pbc unwrap" can not work correctly is simply that the
> atoms move to far between the different frames. If you look at "resid
> 40", you'll notice that in the very first step the molecule moves almost
> half of a simulation box length. In fact, some of the atoms move even
> further. As a consequence, these atoms are wrapped, so that they are
> very far from the other atoms in the same residue.
>
> I don't see a simple way to solve this using the pbctools at the moment.
> One could probably modify the pbctools such that it somehow takes into
> account both the step size (unwrap), the box size (wrap) AND the
> molecular bonds (join), but so far it does only one of it at a time.
>
> I think the only way to circumvent this is to record more frames in the
> trajectory and to make sure that no atom moves further than half a box
> length between too consecutive frames.
>
> Olaf
> 

-- 
Nicola Giacchè, PhD.
Dipartimento di Chimica e Tecnologia del Farmaco.
Faculty of Pharmacy, University of Perugia
via del Liceo 1, 06123 Perugia (Italy)
Tel +39 075 585 5169
E-mail: nicolag_at_chimfarm.unipg.it