From: John Stone (
Date: Wed Oct 02 2019 - 10:08:43 CDT

Josh is correct, you'll either need to select a consistent
set of atoms over the trajectory frames you want to write out,
or you'll need to instead generate completely independent structures
for each trajectory frame, e.g., by writing out a PDB for each
timestep, if you really need to have files for each unique set
of waters.

  John Stone

On Wed, Oct 02, 2019 at 02:43:30PM +0000, Vermaas, Joshua wrote:
> Hi Anupama,
> You can't write out a trajectory for this kind of selection, because the
> number of atoms changes from frame to frame, which breaks typical
> trajectory formats, as well as some assumptions VMD makes internally. If
> there was some core set of water molecules that doesn't change, or if you
> wanted to pick the 100 closest water molecules, some clever tcl scripting
> can do what you want. Would that be helpful?
> -Josh
> On 2019-10-02 07:59:44-06:00 wrote:
> dear all,
> I want to extract the trajectory of water molecules near the protein
> surface within 5 angstorm region. I was able to do that in vmd but it is
> for only one time frame and than the water molecules diffuses away for
> the next time frame. I want to extract the trajectory of only those
> water molecules that are near protein surface within 5 angstorm region
> for each individual time frame so that i can get the complete trajectory
> of water molecules that are near the protein surface excluding the water
> molecules that diffuses away in each next frame......
> I will be thankful for help.
> Thanks and regards
> Anupama,

NIH Center for Macromolecular Modeling and Bioinformatics
Beckman Institute for Advanced Science and Technology
University of Illinois, 405 N. Mathews Ave, Urbana, IL 61801           Phone: 217-244-3349