[gmx-users] Extracting protein plus a shell of water from a .xtc file

[Mark Abraham]

On 16/11/2010 2:30 AM, Paolo Conflitti wrote:
> Dear gmx-users,
>
>  I'm trying to extract a protein plus a shell of water of a given 
> radius from the .xtc and .gro files obtained from a simulation step. 
> My purpose is to make a second box, smaller then the first, with the 
> protein and the shell of solvent in order to optimize the performance 
> of the simulations that I'm running (they require an incredible amount 
> of time due to the high number of H2O molecules in the box) and retain 
> the data obtained till now without starting a new run from the beginning.
>  I've tried to use trjorder with the command -o .pdb -nshell .xvg -r 
> x.x and I've obtained a .pdb file with the distances of the molecules 
> of water from the protein (reference group: Protein, group of 
> molecules to be ordere: SOL) printed in the B-factor field.
>  Now I'm going to modify the pdb file removing all the molecules that 
> are too far from the protein. I will use this file to make a second 
> box, add the water needed to fill it and restart the simulation.
>  I'm concerned about this procedure because I'm afraid that I'll 
> introduce some errors in the new run, so I wonder if there's a faster 
> way to achieve the same result. I've tried different ways, even using 
> other softwares (e.g: MolMol), but I have not been able to extract the 
> shell of water.
>  If there's no other way to do this, could someone please give me a 
> hint about the procedure I'm attempting on?
>  Thanks in advance for your attention and patience.
>

Since you'll have to re-equilibrate this new water box anyway, why not 
just take away all the water, and use editconf and genbox to generate 
the box of the new size? There's no significant advantage to using your 
waters in the .xtc over those generated by genbox.

g_select is a new GROMACS 4.5 tool that is well suited to your approach 
for solving this problem, but I do not think even its limited complexity 
is worth while.

Mark