[gmx-users] PBC problems when creating water layer around protein on an XTC trajectory

[Xavier Periole]

Marc Baaden wrote:

>Hi,
>
>we are currently analysing some trajectories where we would like to look
>at the hydration shell(s) of the simulated protein. We have already treated
>the system with trjconv and trjorder to obtain the centered protein with its
>20000 closest water molecules.
>
>What troubles us is trying to get the "visually" correct water shell, which
>is problematic because the protein is elongated, and using its center of mass
>as reference point (as in -pbc cluster) does not systematically yield the
>right periodic copy of the waters. In some cases one would want the water one
>periodic box above or to the side. In fact to have the right result, several
>periodic copies need to be considered.
>
>The missing option in trjconv that could solve this would be :
>
> -pbc closest_to_any_atom_of_the_reference_group  (eg protein in our case)
>
>because the com of the protein is not the right reference for the waters
>at the edges.
>
>  
>
I am not sure I completely understand your problem, but the solution I 
would give is that
based on the trajectory you already ordered and selected the 20000 
closest waters, you can
get an extimate of the number of water molecules in the first shell 
(using any program giving
you the possibility to select with a distance criterion). Then you can 
generate a index file
that will contain a group defined ans the residue from the first water 
molecule + the number
of molecules. Then use trjconv on the ordered trajectory and print only 
the protein and this
group of N water molecules.

Hope this helps,
XAvier

-- 
----------------------------------
 Xavier Periole - Ph.D.

 Dept. of Biophysical Chemistry / MD Group   
 Univ. of Groningen
 Nijenborgh 4
 9747 AG Groningen
 The Netherlands
 
 Tel: +31-503634329
 Fax: +31-503634800
 email: x.periole at rug.nl
 web-page: http://md.chem.rug.nl/~periole
----------------------------------