VMD-L Mailing List

From: Simon Dürr (simon.duerr_at_uni-konstanz.de)
Date: Mon Sep 21 2015 - 13:19:10 CDT

For those running into this problem.
Try the ILS calculation without the periodic boundary conditions and
see if that works for your system.

I did a pbc wrap and then aligned on the CA-Atoms of frame 0 (using
volutil -alignsel )

In my case it gave very nice results.

Cheers,
Simon

2015-09-11 16:42 GMT-04:00 Simon Dürr <simon.duerr_at_uni-konstanz.de>:
> Hi,
>
> I still have some problems using ILS in VMD. After generating the ILS
> run script with PBC enabled it yields only volumes in one corner of
> the search box. I suspect the alignment to be the problem. In the
> paper (Imaging the Migration Pathways for O2, CO, NO, and Xe Inside
> Myoglobin) and the book chapter (Finding Gas Migration Pathways in
> Proteins Using Implicit Ligand Sampling) it appears that the authors
> are aligning (rot+trans) the trajectory before using the ILS tools.
> However, reading the comments and help texts in volutil it is
> suggested to not align prior to volmap if PBC is used:
>
> “If you use -pbc DO NOT ALIGN the frames of the structure yourself
> prior to the calculation! It will totally mess up the definition of
> your PBC cells. Instead you should use the -alignsel option and let
> volmap handle the alignment ….“
>
> Using the PBC option the generated ILS run script consists of an
> alignment step for the translation only (ILStools::shift_to_center
> $asel) followed by pbc wrap and finally alignsel is used in volmap.
> Could anyone elaborate on the use of ILStools::shift_to_center $asel
> and how it affects alignsel? In addition, should I use an aligned
> trajectory before executing the ILS run script?
>
> In my case the only alignment I see was done by
> ILStools::shift_to_center $asel since all volumes in the DX file are
> in the corner to which my system was aligned by shift_to_center. So
> instead of having the entire system aligned on top of frame 0 the
> system shifts out of the search space.
>
> The simulation I'm using has 10000 frames for 10ns.
>
> Cheers,
> Simon