From: Irene Newhouse (einew_at_hotmail.com)
Date: Tue Aug 12 2014 - 12:45:19 CDT

Averaged structures can result in atoms in non-chemical positions. These indicate regions of high atomic motion during the simulation. If in your averaged structure, these atoms are all hydrogens [I've found that H's are by far the most common cases in my work], you can 'fix' the structure for graphics purposes by removing all the H's and adding them back. If you have sidechain heavy atoms with high motion, you might consider clustering instead of averaging.
 
Irene Newhouse
 
> Date: Tue, 12 Aug 2014 11:54:34 -0500
> From: johns_at_ks.uiuc.edu
> To: kanaan_at_zedat.fu-berlin.de
> CC: vermaas2_at_illinois.edu; vmd-l_at_ks.uiuc.edu
> Subject: Re: vmd-l: averaged structure
>
> I assume ran a periodic simulation, so you're also going to have to deal
> with PBC unit cell coordinate wrapping issues. This is best done by having
> a look at the "pbctools" VMD plugin and doing preprocessing on your structure
> prior to averaging coordinates.
>
> Cheers,
> John Stone
> vmd_at_ks.uiuc.edu
>
> On Tue, Aug 12, 2014 at 06:42:05PM +0200, Natal Kanaan wrote:
> > Thanks for your quick reply Josh & John,
> > I got the averaged.pdb but it is a very weird structure which really can't
> > be used. Do you ever had that problem?
> > Thanks,
> > Natal
> >
> > > Hi Natal,
> > >
> > > Is this running windows? If so, you are probably running into the
> > > limitations of how much memory VMD can address (32-bit applications can
> > > only access 4GB or so of RAM), as your DCDs are probably combined larger
> > > than this limit. There are two solutions to this problem: only take a
> > > subset of the frames (using step as an argument to the addfile command:
> > > http://www.ks.uiuc.edu/Research/vmd/current/ug/node138.html), or go to a
> > > 64-bit system with sufficient RAM to load all the trajectories at once.
> > >
> > > In terms of writing a pdb with the average structure, this should in
> > > principle work:
> > >
> > > set sel [atomselect $mol "whatever text you need to get the selection
> > > you want here"]
> > > $sel set {x y z} [measure avpos $sel]
> > > $sel writepdb averagestructure.pdb
> > >
> > > Be aware that the second line will change atomic coordinates for the
> > > current frame, so instead of it being part of your trajectory, its now
> > > the average position.
> > >
> > > -Josh Vermaas
> > >
> > > On 8/12/14, 7:30 AM, Natal Kanaan wrote:
> > >> Dear all,
> > >> I'm trying to write a pdb structure which is an average of several
> > >> trajectories. Searching in the list, I found a script similar to the one
> > >> below, but I don't know how to really write this new-averaged-pdb
> > >> structure:
> > >>
> > >> set mol [mol new ../model8_final/setup/model8_ionized.psf type psf
> > >> waitfor
> > >> all]
> > >> mol addfile model8_R3_parte1.dcd type dcd waitfor all
> > >> mol addfile model8_R3_parte2.dcd type dcd waitfor all
> > >> mol addfile model8_R3_parte3.dcd type dcd waitfor all
> > >> mol addfile model8_R3_parte4.dcd type dcd waitfor all
> > >> set sel [atomselect $mol all]
> > >> measure avpos $sel
> > >>
> > >> Any kind of help would be really appreciated. Another issue is that, as
> > >> I
> > >> have a lot of data, if I just source this script directly in vmd, it
> > >> ends
> > >> up being "killed". I already tried with:
> > >>
> > >> vmd -dispdev text -e xx.tcl
> > >>
> > >> but it is also killed.
> > >> Thanks a lot in advance for your help,
> > >> Natal
> > >>
> > >
> >
>
> --
> NIH Center for Macromolecular Modeling and Bioinformatics
> Beckman Institute for Advanced Science and Technology
> University of Illinois, 405 N. Mathews Ave, Urbana, IL 61801
> http://www.ks.uiuc.edu/~johns/ Phone: 217-244-3349
> http://www.ks.uiuc.edu/Research/vmd/
>