From: Joshua Adelman (jla65_at_pitt.edu)
Date: Fri Dec 17 2010 - 10:31:41 CST

You don't want to use the 'within' keyword, since that will carve out a strangely shaped box (it's drawing a sphere around each protein atom and only selecting atoms within the sphere). You want something more like:
set sel [atomselect top "{y > -20 and y < 20}]

That will give you a slice of the system that removes extra water above and below your membrane slab.

The other thing you'll have to watch out for is that atomselect will truncate water molecules that straddle the interface of the slab, resulting in a set of atoms that include incomplete water molecules. So you then need to then go back through and get the complete set of atoms using something like:
set B [lsort -unique [$sel get resid]]

and then using that list to get the set of atoms that you want.

Best of luck.

Josh

On Dec 17, 2010, at 9:29 AM, Ajasja Ljubetič wrote:

> I think you have to say
> set selwater [atomselect top "name OH2 and within 10 of protein "]
>
> (*delete the not*)
>
> It's always best to view selections graphically.
> You can put the selection string under representations -> selected atoms and see what you have selected. I usually set the drawing method to CPK.
>
> Regards,
> Ajasja
>
> On Fri, Dec 17, 2010 at 14:18, siddesh southekal <siddeshonline_at_gmail.com> wrote:
> Dear all,
>
> I have a pdb file (after 20 ns MD run ) with a membrane protein (GPCR) embeded in the lipid bilayer with buffer around .There is excess of water molecules in this file . I want to remove this excess buffer such that i have water molecules only 10 A above and below the protein .(say around 15 A around the lipid ) .The membrane normal is in y direction.
> I would be using the truncated file for MD simulation using AMBER.
> I tried -
> set selwater [atomselect top "name OH2 and not within 10 of protein "]
> But whn i try to save the selection the pdb file is blank.How do i exactly do this?
> I would appreciate any suggestions
> Thanks ,
> Siddesh
> Universitaet Bonn
>